Full text of "The prefabrication of houses"

See other formats

THE
PREFABRICATE

HOUSES

G
O

I
I

o"
o

c/o

/' '

THE PREFABRICATION
OF HOUSES

TECHNOLOGY PRESS BOOKS

PRINCIPLES OF ELECTRICAL ENGINEERING SERIES
By Members of the Staff of the
Department of Electrical Engineering,
Massachusetts Institute of Technology

THE MATHEMATICS OF CIRCUIT ANALYSIS

Extensions to the Mathematical Training of Electrical Engineers
By E. A. Guillemin

MAGNETIC CIRCUITS AND TRANSFORMERS

A First Course for Power and Communication Engineers

APPLIED ELECTRONICS

A First Course in Electronics, Electron Tubes, and Associated Circuits

ELECTRIC CIRCUITS

A First Course in Circuit Analysis for Electrical Engineers

THE PREFABRICATION OF HOUSES

By Burnham Kelly

PRESSURES ON WAGE DECISIONS

A Case Study in the Shoe Industry
By George P. Shultz

THE DOLLAR SHORTAGE

By Charles P. Kindleberger

TRANSMISSION OF NERVE IMPULSES AT NEUROEFFECTOR JUNCTIONS AND
PERIPHERAL SYNAPSES
By Arturo Rosenblueth

MID-CENTURY

The Social Implications of Scientific Progress
By John Burchard

AN INDEX OF NOMOGRAMS

Compiled and edited by Douglas P. Adams

EXTRAPOLATION, INTERPOLATION, AND SMOOTHING OF STATIONARY TIME SERIES
with Engineering Applications
By Norbert Wiener

CYBERNETICS

Or Control and Communication in the Animal and the Machine
By Norbert Wiener

Q.E.D., M.I.T. IN WORLD WAR II
By John Burchard

SCIENTIFIC SOCIETIES IN THE UNITED STATES
By R. S. Bates

INDEX FOSSILS OF NORTH AMERICA

By H. W. Shimer and R. R. Shrock

THE MOVEMENT OF FACTORY WORKERS
By C. A. Myers and W. R. Maclaurin

ELLIPTIC CYLINDER AND SPHEROIDAL WAVE FUNCTIONS

By J. A. Stratton, P. M. Morse, L. J. Chu, and R. A. Hutner

WAVELENGTH TABLES

Measured and compiled under the direction of G. L. Harrison

THE

PREFABRICATION
OF

HOUSES

A Study by the Albert Farwell Bemis
Foundation of the Prefabrication Industry
in the United States

By BURNHAM KELLY

PUBLISHED JOINTLY BY
The Technology Press of

The Massachusetts Institute of Technology
AND

John Wiley and Sons, Inc., New York
Chapman & Hall, Ltd., London

BY
THE MASSACHUSETTS INSTITUTE OF TECHNOLOGY

This book or any part thereof must not
be reproduced in any form without
the written permission of the publisher.

PRINTED IN THE UNITED STATES OF AMERICA

'. . . he that strives to touch the starres,
oft stombles at a strawe"

EDMUND SPENSER

The Shepheardes Calender

Foreword

Fourteen years ago Albert Farwell Bemis completed his important
trilogy on housing, published under the general title, The Evolving
House. The final volume of this work, Rational Design, was largely
devoted by Mr. Bemis to exposition of his modular theory, a theory
which has since found wide application in the standardization of the
dimensions of building materials.

The volume also contained a long appendix which I had the privilege
of putting together and which at the time of publication was perhaps
the largest single compilation of the efforts of various people over the
years to arrive at a design for a factory-made house.

This appendix had serious defects, and the greatest of these was
one common to the times, and one from which prefabrication has not
yet escaped, that is, an inordinate interest in the engineering detail of
the various proposals and an inadequate interest in all the other factors
which might determine success or failure.

It is true that I attempted to correct this by publishing a list of
questions which a hopeful prefabricator ought conscientiously to ask
himself, but even these were heavily weighted on the side of design;
and, though the individual descriptions did attempt to state many
facts about each proposal, these facts were obtained from the armchair,
so to speak, by using the replies sent in by the sponsors themselves.
Experience has shown that sponsors are universally overoptimistic.

In the process of putting together this appendix, we naturally ac
cumulated very substantial files of information. Mr. Bemis died in
1936, while his last volume was in the press; in 1938 Mrs. Bemis and
her children established the Albert Farwell Bemis Foundation for
housing research at the Massachusetts Institute of Technology. I had
the honor to be the first Director. The information files came with the
Foundation to M.I.T.

We had scarcely put together a working team when war came along
and scattered it. In 1945 when some of us came back I was soon suc
ceeded as Director by Burnham Kelly, who is the author of this book.

VII

The interest in prefabrication was even more intense in the postwar
period than it had been in the thirties. The files of the Bemis Founda
tion, though far from complete, were certainly among the largest in
the country. The Foundation was frequently sought out by visitors,
especially from abroad, who were seeking the truth about a business
concerning which many half-truths or untruths were being said. In
the light of this interest it soon became apparent that we needed much
more information than we had, and of many different kinds, if an
approximation to truth was going to be possible. It was also clear that
much of this information could be obtained only by personal observa
tion in the communities of the various entrepreneurs. It was this that
started the Foundation to collecting more information by the process
of field survey. It is the results of this field survey and the conclusions
which may be drawn from them that the reader will find in this book.

Prefabrication, or the factory manufacture of houses, means many
different things to different people. To some it is a variegated Gol-
conda; the seeker for a house who finds that what he does not want
costs more than what he wants to pay imagines that houses produced
like automobiles or radios ought to be nearer his heart's desire; the
entrepreneur imagines that he may be another Ford; the manufacturer
of conventional building materials wonders whether he may not sell
more of these by making them into some sort of package; the manu
facturer of building equipment fancies that he may have all his latest
apparatus in every house if he provides the package as well; a national
president faced with depression may look to it as a new industry to
lead from the morass; the opponent of subsidized housing may see a
chance of arresting the tide if the cost of the housing unit can be
materially reduced through factory methods. And all these hopes
would have some justification if only the successful commercial manu
facture of houses on a large scale could be achieved.

To others prefabrication is a source of fear and not of hope. The
investor who is overcommitted in loans on real estate may legitimately
wonder whether a sudden and significant downward shift in the cost
of a house may not be disastrous; the building-trades laborer who
pursues an antiquated craft with little of the joy of the onetime build
ing craftsman may fear technological unemployment; the realtor who
is not wise about real estate and is really nothing more than a peddler
of some one else's property may have the same apprehension; and to
a certain extent every present homeowner can share the fear of the
investor although he may display it in an attitude towards the appear
ance of the product. All these vested interests are precisely the same
in kind as those which have historically opposed every other innova-

viii

tion (and which in the long run have always lost out). And the
methods they use to oppose are fundamentally the same-the marshal
ling of adverse public opinion, the imposition of restrictive legislation,
the technique of the strike. All their fears would have some justifica
tion if successful commercial manufacture on a large scale came about
too suddenly.

But there are others, too, who are interested. There are sincere en
gineers and inventors who think that by application of their personal
talents something socially important (and personally profitable) will
come about. There are blageurs who are more interested in personal
publicity than in a successful house and who, therefore, propose pre
posterous but fascinating fantasies. These take the eyes of publishers
who have magazines rather than realities to retail, and they serve as
interesting table conversation among the avant garde; unfortunately,
they also raise hopes, only to shatter them again. This has been going
on for a long time— too long.

It is of course always possible that some miraculous invention may
open the gates which have so long resisted all attack, but this seems
very unlikely. It has seemed to some that enormous investments of
capital might offer the key, but that this in itself is not enough seems
witnessed by some recent events. Some of us have hoped that a thriv
ing, if small and unspectacular, manufacturer of fairly conventional
houses, might, step by step, year by year, introduce the improvements
in structure and materials and the efficiencies in design and production
which would gradually drive down the cost and increase the market.
But none of these things has yet come about.

I have spent so much discussion on what is after all a foreword
largely as an easy way of saying what was wrong with my appendix
of 1936 and of appreciating the corrections which I believe the reader
will find in the present volume. It should be of interest to all those I
have cited as having hopes or fears, to houseseeker and houseowner,
to investor and realtor, to manufacturer looking hopefully beyond his
own range, to building-trades laborer, to politician, and to statesman.
At the most it may suggest how really to open the gate; at the least
it will suggest how not to.

Successful factory manufacture of houses will depend upon a first-
rate combination of managerial brains, financial acumen, engineering
skill, aesthetic sensibility, social consciousness, and marketing wisdom.
A study of the state of the art stands therefore at the crossroads of the
applied physical and social sciences, an appropriate place for a teacher
at M.I.T. to stand.

It is therefore a pleasure to say of this book by Mr. Kelly that I be
lieve it has an important message to tell. I am gratified that he has
done such a good job; and I think that Mr. Bemis, if he were alive,
would be gratified, too.

JOHN ELY BURCHARD
Cambridge, Massachusetts
January, 1951

Preface

In the hope of serving readers of widely varying interests, this book
has been divided into three parts. The first part is editorial in nature.
It includes a brief history of the prefabrication of houses in the United
States, a summary of the present state of the industry, and some specu
lation regarding its future. This material represents the best judg
ment of the Bemis Foundation, and, while it has a broad basis in fact,
we have not hesitated to generalize, to extrapolate, and to present our
unsupported opinions.

The second part of the book can better be described as reportorial.
Here the industry has been treated as nearly as possible on a factual
basis, with opinions given only when there is no other way by which a
trained observer may record the facts. This material approaches the
status of working data, and we hope that it may be used by men wiser
than we to correct the conclusions reached in the first part.

The third part of the book is a collection of more detailed appendix
material, not suitable for inclusion in the text, but likely to prove use
ful to many readers.

We have emphasized throughout the book the importance of treat
ing the prefabrication of houses as a complete pattern of operations of
which management, design, procurement, production, and marketing
are the major subdivisions. Indeed, the material in the second part is
so organized that the reader will have to look under each of these
subdivisions in order to gather all the information on any one com
pany. To have organized the material by companies, while maintain
ing this emphasis, would have meant endless repetition and a doubling
of the bulk of the book.

Far more important, the full understanding of each subdivision of
the pattern of operations might have been lost in a company-by-
company analysis. We hope to make it abundantly clear that the
company which has good design must also have good management,
intelligent procurement, efficient production, and effective marketing
to have any chance of real success.

A large part of the material on which the book is based was collected
over a period of years in the files of the Bemis Foundation and its
antecedent, Bemis Industries, Inc. The backbone of this study, how
ever, was a detailed survey of prefabricators in the United States which
was made by the Foundation in 1946 and 1947 and supplemented by
an extended field survey by our Research Assistant, Herbert S. Heaven-
rich, Jr. At that time, there were in the field more prefabricators and
would-be-prefabricators than there had been ever before or have been
since. With a postwar housing boom in view, and with the encourage
ment of government and financial circles, many of those whose in
genuity and productive skill had proved valuable in the war effort
determined to invade the field of housing. The high noon of this
effort occurred, by chance, just at the time of our survey. Through
good fortune, therefore, we are able to present an analysis of some
general historical value.

We are also fortunate in that relatively few new developments of
importance have taken place during the period required for the
analysis of our information and the preparation of this book. Recent
news in prefabrication has consisted largely of the failures of some
companies and the continuing development of others whose character
was already well established in 1947.

The work of writing this book was shared in large part by the sev
eral Research Assistants of the Bemis Foundation.

The original organization of the field survey and the first assembly
of the material in the second part were the work of Herbert S. Heaven-
rich, Jr. William F. Blitzer helped to develop the final form of the
book and wrote drafts of many major sections of it, including in the
first part the chapters on the history and present state of the industry,
and in the second part the chapters on management, procurement, and
production. Cyril C. Hermann put together the material on marketing,
and John F. Falkenberg and Barbara W. Atchley assisted in the final
editing process.

Many of the concepts which are developed in this book are those
of John E. Burchard, who was Director of the Bemis Foundation when
this study was started, and whose special knowledge and experience
in the field have been of great value to us. Although we have made
many references to his writings, it would not be possible by such means
to acknowledge the degree to which we have benefited from his in
sight.

Our debt to those working in the field, whether as actual producers
of houses or in collateral positions, will be evident throughout the

fctf

book. We should like to express our deepest gratitude here, however,
for the friendly cooperation and intelligent criticism which they have
offered us from the start.

BURNHAM KELLY

Cambridge, Massachusetts
January, 1951

xttt

Contents

PART I PAGE

Chapter 1 DEFINITIONS 1

Chapter 2 THE DEVELOPMENT OF THE INDUSTRY 5

I. Before 1900: Beginnings 7

A. First Traces in America 7

B. Early Prefabrication in Cast Iron 8

C. Impetus of the Gold Rush 9

D. Prefabricated Camp Buildings 9

II. 1900-1920: Developments in Precut and Concrete

Construction 11

A. The Precut House 11

B. Early American Experimentation 12

C. The Emphasis on Concrete 14

III. 1920-1930: Experimentation with Prefabrication 15

A. The Postwar Stimulus Abroad 15

B. Experimentation and Small-Scale Development in

America 20

IV. 1930-1940: Prefabrication Attains the Status of a

Movement 28

A. The Background Influences 28

B. Non-Commercial Research and Development 31

C. Government Activity: Techniques and Standards 33

D. Government Activity: Prefabricated Construction 35

E. Commercial Development by Private Enterprise 38

F. General Trends and Characteristics 46

G. The Analogy with the Automobile 51

V. 1940-1945: The War Period 55

A. Prefabrication on Trial

B. Factors Favorable to Prefabrication 57

C. Signs of Prefabrication's Growth 58

D. The Contribution of Prefabrication 60

E. The Effect of the War on Prefabrication 62

PAGE

Chapter 3 1946-1949: GREAT EXPECTATIONS AND DISAP
POINTMENTS 65

I. Background 67

A. The Shortage 67

B. The Wyatt Program 68

C. The Birth and Death of Firms 71

D. The Building Boom 72

II. The Prefabricator: A Stage in Industrialization 74

A. The Panelized Wood Frame House 74

B. The Stressed Skin Plywood House 77

C. The Machine-Made Metal House 79

D. Other Types of Prefabrication 80

III. Broad Aspects of Prefabrication 81

A. Modular Coordination 81

B. The Rationalization of On-Site Building 83

IV. Prefabrication: Nature and Cost of the Product 84
V. Prefabrication: Current Problems 86

A. Locus of Operations 86

B. Marketing 87

C. Public Acceptance 89

D. Building Codes 90

E. Local Trade and Labor 92

F. Financing 92

VI. Conclusion 95

Chapter 4 THE FUTURE OF PREFABRICATION 97

I. Introduction 99

II. Current Trends within the Industry 99

A. Management 100

B. Design 101

C. Procurement 106

D. Production 108

E. Marketing 110

III. Future Problems within the Industry 116

A. Central or Branch Plants 117

B. Site or Factory Fabrication 117

C. Low Price or High Value 118

D. Evolution or Revolution 119

E. One Model or Many 120

F. Optimum Level of Standardization 121

G. Duplication by the Conventional Builder 121

IV. Larger Housing Issues 122
A. The House Itself 122

xvi

PAGE

B. The Community 128

C. Broad Economic and Policy Problems 131

V. Conclusion

135

PART II

Chapter 5 INTRODUCTION 139

Chapter 6 MANAGEMENT 143

I. Background 145

II. Labor Relations 148

Labor Relations in the Plant 148

A. Unions 148

B. Wages 150

C. Restrictive Practices 153
Labor Relations in the Field 155

III. Financing 157

A. Capitalization 157

B. Sources of Investment Capital 159

C. Credit 163

IV. Public Relations 168
V. Trade Associations 170

A. Prefabricated Home Manufacturers' Institute 170

B. National Association of Housing Manufacturers 172

Chapter 7 DESIGN 175

I. Introduction 177

II. Classification of Prefabrication Systems 180

A. By Materials 180

B. By Structural System 185

C. Miscellaneous Classifications 188

III. Description of Components 196

A. General 196

B. Foundations 198

C. Floors 202

D. Walls 210

E. Ceilings 244

F. Roofs 253

IV. Miscellaneous Design Features 262

A. Plumbing 262

B. Mechanical Cores 265

C. Heating 268

D. Electrical Wiring and Fixtures 273

xvii

PAGE

E. Acoustical Treatment 274

F. Built-in Furniture 274

G. Space Arrangement 275
H. Product Variety 281

Chapter 8 PROCUREMENT 287

I. Raw Materials 289

II. Finished Material and Equipment 293

III. Fabricated Components 297

Chapter 9 PRODUCTION 301

I. Plant Facilities 303

II. Location of the Industry 304

III. Labor Force 308

IV. Factory Processes and Equipment 309

A. Wood 309

B. Metal 318

C. Concrete 321

D. Honeycomb Core Sandwich Materials 325

V. Some Particular Aspects of Production 327

A. Factory Storage Facilities 327

B. Plant Layout 328

C. Production Scheduling 331

VI. Analysis 333

A. The Amount of Manufacture by the Prefabricator 333

B. Production Volume and Production Costs 338

C. Productivity 340

D. Production Costs 344

Chapter 10 MARKETING 357

I. Introduction 359

II. Markets 360

A. Market Areas 360

B. Special Market Types 364

III. Pricing Policies 367

IV. Channels of Distribution 372

A. Factory Direct to Consumer 373

B. Factory to Dealer to Consumer 376

C. Factory to Distributor to Dealer to Consumer 382

V. Sales Methods 385
VI. Financing the Prefabricated Home 391

A. Financing the Dealer 392

B. Financing the Purchaser 394

xviii

VII. Choosing the Site

VIII. Transportation to the Site

IX. Erection of Prefabricated Houses

X. Service to Customers after Erection

XI. A Review of Failures

Chapter 11 CONCLUSION
Appendix A PROCEDURE

Appendix B

Appendix C
Appendix D

PAGE
398

400
407
411
413

417

423

COMPANIES AND PEOPLE VISITED 438

PREFABRICATORS MENTIONED BUT NOT VISITED 442

LISTS OF PREFABRICATORS

ANNOTATED BIBLIOGRAPHY
I. Books and Pamphlets
II. Conference Proceedings

III. Trade Association Material

IV. Periodicals

V. Other Sources

INDEX

444

446
446
449
450
451
451

453

XIX

List of Illustrations

PAGE

1. Skillings patent drawings 10

2. Precast concrete systems of the 1920's 22 23

3. A typical wood frame panel 75

4. A typical stressed skin panel 78

5. The principles of modular coordination 82

6. Hodgson houses following 96

7. A precut house of 1920 " 96

8. Buckminster Fuller's first Dymaxion house " 96

9. Early General Houses house " 96

10. Three prefabricated houses of the 1940's " 96

11. Two circular houses " 96

12. A folding unit designed for emergency shelter " 96

13. Lustron houses " 96

14. Conventional framing illustrating construction terminology 186

15. The General Panel system 192

16. The Acorn footing 201

17. Examples of grade beam and concrete slab on grade 209

18. Commonly used panel joints 220

19. Metal construction systems 224, 225

20. Sandwich panel materials 234

21. Southern California Homes house following 256

22. Reliance house 256

23. Section of AIROH house " 256

24. Pierce Foundation— Cemesto House 256

25. Production Line Structures 256

26. Wingfoot house 256

27. Acorn house 256

28. The Fuller house 256

29. Kaiser Community Homes house 256

30. Green's "solar house" 256

31. Patent drawing for the integrated Fuller bathroom 264

32. Exploded drawing of Ingersoll Utility Unit 266

33. Pkns of selected prefabricated homes 276-279

34. The location of prefabrication plants in 1948 305

35. Ford house following 320

36. Butler house

37. LeTourneau system

38. Ibec system 32°

xxi

PAGE

39. Gunnison plant operations " 320

40. National Homes plant operations 320

41. Lustron plant operations " 320

42. Reliance plant operations " 320

43. Crawford Corporation " 320

44. Texas Housing Co. " 320

45. Lustron plant flow diagram 330

46. National Homes thrift model following 416

47. Two other economy models 416

48. National Homes house being erected " 416

49. Lustron house en route 416

50. National Homes house en route " 416

51. The development of a National Homes project 416

52. Lustron house being erected " 416

xxn

Part

Chapter

DEFINITIONS

The term "prefabrication" has often been loosely applied to any
type of novel construction or to any method of building which differs
in some significant respect from conventional construction. This
stems from the plain truth that it is a difficult term to define, as can be
shown by a consideration of some of the definitions which have been
attempted.

One of the most general definitions, and one of the most official, is
the following:

A prefabricated home is one having walls, partitions, floors, ceilings,
and/or roof composed of sections or panels varying in size which have
been fabricated in a factory prior to erection on the building foundation.
This is in contrast to the conventionally built home which is constructed
piece by piece on the site.1

Other writers try to be more specific:

... It is a question of degree. To oversimplify, and to look a bit into
the future, if you shove and snap a product into place in the field, that is
prefabrication. If you mix, cut, spread, fit, and patch— that's not pre-
fabrication. If the field operation is essentially assembly, rather than manu
facture, you have prefabrication. A brick and plaster wall, of course, em
ploys manufactured ingredients, but such a wall is really manufactured in
the field.

The amount of scrap and waste that must be cleaned up and removed
from a building site may be taken as a rough index of the degree of pre
fabrication employed in any given building operation, since waste results
principally from a manufacturing process, not an assembly process.2

Or are content to be more general:

... a movement to simplify construction by increasing the proportion of
work completed before erection.3

And a few have been driven to extreme conclusions:
Prefabrication is a state of mind.4

1 Prefabricated Homes, Commercial Standard CS 125-47, 2nd ed., Prefabricated
Home Manufacturers' Institute and U. S. Department of Commerce (Washing
ton, 1947), p. 1.

2 Howard T. Fisher, "Prefabrication. What Does It Mean to the Architect?"
Journal of The American Institute of Architects, X (November 1948), 220.

3 Quoted in Proceedings, American-Soviet Building Conference (held under
the auspices of the Architects Committee of the National Committee of Ameri
can-Soviet Friendship in cooperation with the New York Chapter of the Ameri
can Institute of Architects; published in collaboration with The Architectural
Forum, 1945), p. 43.

4 Robert W. McLaughlin, quoted in Proceedings, American-Soviet Building
Conference, p. 43.

From these definitions it can be seen that, in general usage,
"prefabricated" construction is "unconventional," but not all "uncc
ventionar construction is "prefabricated." Secondly, there may
said to be various degrees of prefabrication, of which precutting mig
be one, the fabrication of panels another, the construction of volun
enclosing sections a third, and the manufacture of a complete mob
dwelling unit probably the ultimate.

For our part, we shall attempt no general definitions of the ter
Indeed, it is a major argument of this book that the distinction I
tween prefabricated and conventional construction may well becoi
meaningless within the next few decades. Nevertheless, it is tr
that this study is concerned primarily with those companies whi
are organized to manufacture and in some degree to assemble off t
site one or more of the basic components of a house, such as founc
tions, floors, walls and partitions, ceilings, and roofs.

Some of the houses produced in this manner are completely conve
tional in final construction; it is only the process by which they 2
manufactured and assembled which distinguishes them from t
product of the local builder. A large operative builder, develop!
raw land, making bulk purchases of materials and equipment, a:
building 200 or more houses at one time with work crews which mo
from house to house performing highly specialized functions, may
the end offer for sale houses which are far less conventional than the
of a prefabricator.

In this study, the operative builder is distinguished from the pi
fabricator because his houses are manufactured and assembled large
on the site, but attention is called to him because of the efficiency
some of his methods. In the long run he is to be considered not
much a competitor of the prefabricator as a potentially good custom

Part A •

£*

Chapter

THE DEVELOPMENT

OF THE INDUSTRY

This historical study is concerned with prefabrication— and almost
exclusively with prefabrication in the United States as an industry,
rather than as a trend or movement. It is written from the point of
view of economic history rather than of technical history; that is, pre
fabrication is treated more as an industrial development than as a
succession of ideas about design. The latter subject has been very
well covered elsewhere.1 This study will concern itself with the ex
tent to which these technical ideas were realized in production and
the factors which led to their abandonment or adoption. Thus it is
not so much a discussion of invention— the disclosure of a new method
of achieving some technical objective— as it is of innovation— the com
mercial introduction of a new or improved product or process.

I. Before 1900: Beginnings

A. First Traces in America

A search for the earliest historical evidences of prefabrication would
lead us back to the burnt clay bricks of the Mesopotamian civiliza
tions, many centuries before the Christian era— perhaps further. For
our purposes, however, it will suffice to know that as early as 1624
the English brought with them to Cape Ann a panelized house of
wood for use by the fishing fleet and that this house was subsequently
disassembled, moved, and reassembled several times.2 In 1727 two

1 Albert Farwell Bemis and John Burchard, 2nd, The Evolving House: Vol. Ill,
Rational Design (Cambridge: The Technology Press, 1936).

Alfred Bruce and Harold Sandbank, A History of Prefabrication (Raritan,
N. J.: The John B. Pierce Foundation, 1943).

Housing Production II, or The Application of Quantity Production Technique
to Building: Some Technical History and Considerations, Second Report of the
Committee for the Industrial and Scientific Provision of Housing (London, 1943).

D. Dex Harrison, J. M. Albery, M. W. Whiting, A Survey of Prefabrication
([London]: Ministry of Works, 1945). This survey of prefabrication designs is
among the .most complete of its type. Frequent reference to it is made in this
book.

2 Charles E. Peterson, "Early American Prefabrication/' Gazette des Beaux-
Arts, XXXIII (January 1948), 38.

houses " 'all cut to be erected' " 3 were exported from New Orleans to
the West Indies, and there are other signs of the use of prefabrication
throughout the next 100 years as a means of providing persons with
shelter immediately on their arrival at a new settlement.4

B. Early Prefabrication in Cast Iron

Perhaps the earliest metal prefabricated house was that built some
time before 1830 for the lockkeeper at Tipton Green, Staffordshire,
England. The walls were of flanged vertical cast-iron panels, bolted
together, painted on the exterior and lathed and plastered on the in
terior.5 Cast iron was prominent in a number of other early experi
ments in prefabrication. Watt and Boulton in England began erect
ing their cast-iron framed factory buildings in 180 1,6 and in America
during the forties and fifties cast-iron columns and repetitive elements
of cast iron and glass were used, respectively, for frame and en
closure, as seen notably in the fagades of the commercial buildings
by James Bogardus.7 The use of cast iron in prefabrication reached
a spectacular climax in the Crystal Palace, built in Hyde Park, London,
in 1851. Joseph Paxton based the design of this building on a rela
tively few mass-produced elements: glass panes, wood frames in
which these were set, and cast- and wrought-iron columns and girders
which were bolted together at the site to form the framework.8 Pro
claimed the largest single building the world had yet seen, it was
erected in a mere four months,9 and, demountable, it was later moved
to Sydenham where it was re-erected. The Crystal Palace was a
marvel for the light and airy quality of its structure, in some respects
a reflection of Paxton's experience with greenhouses, but more than
this it was a daring adventure in the use of carefully designed fac
tory-fabricated components and of precision rather than sheer mass
to achieve structural strength.

8 Loc. cit.

* Loc. cit.

5 Harrison et al., op. cit., p. 3.

6Sigfried Giedion, Space, Time and Architecture (Cambridge: Harvard Uni
versity Press, 1941), pp. 124-7.

* Ibid., pp. 129-34.

8 Giedion, op. cit., p. 186.

9Tallis's History and Description of the Crystal Palace, ed. by J. G. Strutt,
in 3 vols. (London: The London Printing and Publishing Co. [c. 1851]), Vol. I,
p. 11.

C. Impetus of the Gold Rush

These, however, were but sporadic beginnings. The first real im
petus to the production of prefabricated houses appears to have
been the Gold Rush of 1848. 10 Houses were exported to California
from our eastern seaboard, from England, France, Germany, Belgium
—even from China, New Zealand, and Tasmania. In the New York
area alone some 5,000 houses for shipment to California had been
contracted for or produced by 1850. Models which cost $400 in the
East sold for $5,000 on the West Coast. And from Manchester, Eng
land, came several hundred houses of corrugated galvanized iron,
some of them outfitted with wallpaper, carpets, furniture, and water
closets. But the end of this boom— which was the first of several
which were to push prefabrication for one reason or another— came in
1850, when the building materials market in California was flooded
and prices fell sharply. After a local lumber industry had developed
in response to the huge demand, the high shipping cost quickly put
the imported prefabricated house out of the picture.

D. Prefabricated Camp Buildings

New settlements provided one market for early prefabricators; the
demand for various types of camp buildings and cottages provided
another. The continuing commercial development of panelized wood
houses for this market dates from at least as far back as 1861. In that
year Skillings and Flint, lumber dealers of Boston and New York,
patented a system of building houses from a few standardized panels
and a number of other interchangeable parts (see Figure 1). Their
impressively bound, gilt-edged catalogue claimed that their houses
could be erected in three hours and showed a number of designs
suited to plantation and army camps. Indeed, this firm sold a good
many houses to the Union Army.11 In Germany the firm of Christoph
& Unmack, organized in 1882, was soon to begin production of timber
houses constructed of load-bearing panels. Its products were chiefly
huts, cabins, and labor camps and were ultimately to be shipped in

10 A very interesting account of this boom in prefabrication is given by Peter
son, op. cit., pp. 42-6.

11 Peterson, op. cit., p. 46.

Figure 1. Shillings Patent Drawings

very large quantities to many parts of the globe.12 A decade later, in
1892, Ernest F. Hodgson founded his company in Boston and began
the manufacture of panelized dwellings of wood ( see Figure 6 ) . This
firm, which is the oldest known to be still in the industry today,13
started by making small structures such as chicken houses, children's
play houses, and dog houses. It received a boost when, with the ad
vent of the horseless carriage, the demand for "auto stables" arose.
A further effect of the automobile was to increase the demand for
vacation cottages, enabling the company to enlarge its operations and
furnishing a major part of its business until this day. This enterprise
has been a notable instance of successful commercial development
on a conservative basis. The house remained of essentially conven
tional wood frame construction, modified to permit shop fabrication
in panels. There was no rush to get into large-scale production nor
any attempt to provide a universal set of building components adapt
able to any plan. Emphasis was placed on modest single-story houses,
and sales were direct to the customer, featuring speedy erection and
good quality rather than low cost.

II. 1900-1920: Developments in Precut and Concrete Con
struction

A. The Precut House

Shortly after 1900 a peripheral development of prefabrication be
came commercially important. This was the precut house, some
times referred to as the "mail-order" house, and in some respects the
first of the "self-help" houses designed for owner erection (see Fig
ure 7). The first decade of the century saw the entrance into this
field of a number of firms which were to become prominent in it:
The Aladdin Co., Bay City, Mich.; Gordon-Van Tine Company, Dav
enport, la.; Pacific Systems Homes, Inc., Los Angeles, Calif.; Sears,

12 Harrison et al., op. cit., case sheet on Christoph & Unmack.

13 The E. F. Hodgson Co. was combined with Allied Housing Associates, Inc.,
in 1944 to form the Allied-Hodgson Housing Corp. Each of the original com
panies does business as a division of the corporation.

Roebuck and Co., Newark, N. J. Although the mere precutting,
notching, and marking of the lumber to be used in a wood frame
dwelling might not appear to deserve the name "prefabrication," the
precut house warrants consideration here for several reasons. One is
that the grading, cutting, marking, and packaging of lumber and the
preassembly of windows and doors in the precutter's plant have usu
ally been carried out on an efficient line production basis. Secondly,
the precut house has generally involved a number of standardized
products in a field that long resisted standardization. Thirdly, it has
made possible the large-scale estimating, purchasing, and collecting
of materials (including roofing, shingles, hardware, etc.) to form a
house package and has established the fixed price character of this
package. Lastly, precutting was, until World War II, probably the
most widely used application of factory production to housing; "cer
tainly a quarter of a million houses have been built according to this
method— a number probably in excess of the total number of sectional
and prefabricated houses built to date (1943), including the war
time demountables." 14

B. Early American Experimentation

It was also about the turn of the century that early experimentation
began in America; it was concerned primarily with concrete as a mate
rial, and here we may mention Grosvenor Atterbury as a pioneer.
His research in the techniques of housebuilding began in 1902, first
at his own expense, subsequently with philanthropic support (after
1907 chiefly from the Russell Sage Foundation). He has continued
until the present day his search for better methods of construction
with cast cementitious materials. About 1907 Atterbury developed
a system of large15 precast hollow-core panels for walls, floors, and
roofs. Between 1910 and 1918 several hundred houses based on this
system were built for the Russell Sage Foundation in Forest Hills,
Long Island, the units being transported to the site in trucks and
erected there with derricks. This was a significant experiment in a
new construction technique, yet the importance of Atterbury 's work
lies not so much in the achievement at Forest Hills, which, though of
high quality, was not of radically low cost, but rather in his approach

14 Bruce and Sandbank, op. cit., p. 57.

15 Wall panels, for instance, were of story height and 6' 0" to 8' 0" wide.

and in the persistence which marked his attempt to develop some sort
of casting process, a method for which he saw great promise in build
ing. His system— along with others involving large precast concrete
units— entailed difficulties in the transportation and handling of heavy
units and in the large investment in molds which was required be
cause of the lengthy curing period for each casting. Therefore, al
though a large-scale project such as that at Forest Hills might over
come these difficulties economically, it was evident that the system
was not well suited to the erection of isolated free-standing houses or
small developments. To solve these difficulties, Atterbury has since
experimented with various other cementitious materials and has de
veloped better molds and worked out methods of shortening the cur
ing time.

Another development in which Atterbury pioneered was the growth
of interest in prefabrication as a means of providing shelter for the
vast bulk of our housing needs, not for just a few exceptional ones.
Before the early twentieth century, the prefabricated dwelling had
been of importance for new settlements, camp cabins, and vacation
cottages— uses in which a decrease in site work was desired even, if
necessary, at an increase in total cost. But as the ever-accelerating
industrialization of our life proceeded along with a great surge of
urbanization, and as our attitudes towards slums and blight changed,16
it was felt that house production methods were falling far behind
industrial techniques in other fields, and prefabrication came to be
regarded as a means of providing more economic shelter for the
mass of our housing requirements. There was no overnight change,
of course, nor can the turning point be placed with too much certainty
in the first decade of the century.17 Yet it seems reasonably clear
that after this time the interest in prefabrication was connected less
and less with a few special housing markets and more and more with
low cost in the general housing market; it was increasingly an interest
in the overall industrialization of house production as an answer to
what was gradually to become known as "the housing problem."

16 Evidenced by the passage, between 1870 and 1890, of many municipal ordi
nances governing the health and safety standards of housing.

17 Atterbury 's work is, however, the earliest example our research has dis
closed of philanthropically supported experimentation in prefabrication; the re
sults, including patents, were offered to any "non-profit institution willing to
continue the work along proper lines looking towards a scientific solution of the
housing problem." (Quoted from a statement by Atterbury submitted to A. F.
Bemis in 1935, files of the Bemis Foundation.)

C. The Emphasis on Concrete

The work of Atterbury was by no means the only experimentation
with novel housebuilding methods during the first two decades of the
century.18 Many of the other attempts also employed concrete and
generally involved but little prefabrication; they were limited to the
precasting of wall units or the use of factory-produced forms. In
1908 Thomas Edison proposed a method of casting two- and three-
story houses in one operation. Sectional cast-iron forms were to be
bolted together at the site, and concrete, carried by a conveyer, was
to be poured into a funnel at the top of the enclosure. Edison's idea
attracted a good deal of attention but was soon abandoned as im
practical. Yet it is interesting to note that the monolithic concrete
house is still a subject of considerable interest and is today being car
ried out in single-story dwellings with equipment at least as complex
as Edison proposed.19 Other ideas were to follow: the Merrill Sys
tem of monolithic concrete walls formed in situ (1908); Simpson
Craft, a complete house system of concrete, about 90% precast ( 1917 ) ;
Lakeolith, the precast ribbed panel system of Simon Lake, the sub
marine designer (1918); the Hahn Concrete Lumber System of pre
cast and site-formed concrete (1919).20 Some hundreds, perhaps a
few thousand, houses have been produced by these and similar
concrete constructions, but no one system has ever been adopted on a
mass-production basis. The early experimental work in concrete
did not develop any fully realized techniques; it was rather a sign-
perhaps the first sign— of the growing interest in the invention of pre
fabrication systems; it was in a sense the forerunner of what we call
the prefabrication movement.21

18 Bemis and Burchard, op. cit.t p. 61T, list nine other examples.

19 For example, R. G. LeTourneau Inc/s Tournalayer and Ibec Housing Cor
poration's house-sized form.

20 For further information on these and other systems see Report on Survey of
Concrete House Construction Systems, Portland Cement Association (Chicago,
1934); also works cited in footnote, p. 7.

21 Bruce and Sandbank, op. cit., pp. 30-40.

III. 1920-1930: Experimentation with Prefabrication

A. The Postwar Stimulus Abroad

We have passed over the World War I period because, unlike the
recent war years, it was not very important to prefabrication. The
production of precut and to some extent of panelized wood buildings
was stimulated, but prefabrication as an industrial development was
not appreciably promoted or advanced. The postwar years, however,
did bring a strong stimulus to prefabrication, chiefly in Europe.
While America continued to experiment with prefabrication, Europe,
by contrast, built with it, and we might digress for a moment to
consider what was accomplished there and why.

1. Great Britain

The earliest developments were in Great Britain, where the hous
ing shortage, the dearth of bricks and of bricklayers and other crafts
men, and the surplus of steel capacity all combined to provide a strong
economic motivation for trying new methods of building. Most of
the British preferred brick, but alternative constructions had to be
available in case of trouble, so that the government could perform
on its promise of "homes for heroes."

By 1920, the Ministry of Works had approved some 110 systems
of construction, of which, excluding systems of concrete masonry,
perhaps 12 involved some degree of prefabrication, though not even
all these reached the production stage.22 There were no standards
of functional performance employed, nor were the systems approved
necessarily cheap or easy to erect. Rather, the emphasis was on
meeting the situation described above, and so, between the years
1918 and 1925, a large number of partially prefabricated houses were
built of elements such as sheet steel, rolled steel frames, concrete
masonry, story-height precast concrete units, and expanded metal
sprayed with cement. The last type of construction, combined with

22 Harrison et al., op. cit., p. 5.

a steel frame, formed the basis of the Dorlonco house, some 10,000
of which were built in England between 1920 and 1928.23 Many later
proved defective in that, because of insufficient cover, the metal lath
rusted and the cement rendering cracked and fell off. In the years
following 1918 some 10,000 concrete houses were erected by four big
industrial concerns using either precast pier and panel construction
or precast slabs to enclose a site-poured frame, and from 1926 to
1928 another big corporation, G. and J. Weir, Ltd., built 3,000 houses,
using timber frame, steel plates, and fiberboard— materials relatively
rare in British housing. It is important to note that until World War
II, with the exception of the precut and possibly certain panelized,
but otherwise conventional, wood frame houses, no American pre-
fabrication system was produced as extensively as the Weir and Dor
lonco houses.

Yet even with such extensive trials, all these prefabrication systems
fell into general disuse between 1926 and 1930. For one thing, there
was labor trouble in connection with the Weir houses because of the
fact that engineering union labor was used at lower wage rates than
unskilled building-trades labor; as a result the building trades refused
to work for any local authority which erected Weir houses. But,
more than this, it is important to understand that the prefabricated
houses were considered as makeshifts and, at least in part, as pawns
in the struggle with conventional building labor. When the shortage
was overcome and the normal building methods could handle the de
mand, few prefabricated houses were built. They had not proved
cheaper than the brick houses in whose stead they were being built,
and they could not compete on even terms because of the prejudice
against them as being new, untried, and substitute products.

. . . the new types were, in design, mostly inferior imitations of brick
buildings. No attempt had been made to evolve designs which suited,
and took advantage of, the new structural concepts. So utterly bankrupt
was the movement in this respect that the new constructions were labori
ously worked to the same niggling plans which were in common use for
brick houses at the same time. It was not realized, and it is still not
realized, that plans and designs suitable for brick buildings, which can be
cut and chopped about in extraordinary detail, are totally unsuited to the
factory-made articles of standardized size which require the clearest and
simplest planning for their economical use.24

This statement might well form the epilogue to a number of other
essays in prefabrication.

2*Ibid., case sheet on Dorlonco.

24 D. Dex Harrison, "An Outline of Prefabrication," in Tomorrow's Houses,
ed. by John Madge (London: Pilot Press, 1946), pp. 118-9.

2. Germany

The Germans, because of the economic consequences of the war,
did not begin to experiment with prefabrication on a large scale until
about 1926, by which time the British were already returning to con
ventional building methods. As in Britain, however, the necessity
of providing housing under abnormal conditions brought into effect
many new approaches. A large experimental program, exceeding in
scale anything attempted previously in any country, was carried out
under the direction of the State Research Institute ( Reichsforschungs-
gesellschaf t ) , a government department charged, among other duties,
with organizing and controlling building throughout Germany. Many
new schemes for low-cost housing construction were tried, and the
costs and physical results compared. A big slump in the steel trade
in 1927 left Vereinigte Stahlwerke, the German steel trust, with con
siderable excess capacity and the desire to seek new outlets. As a
consequence, this trust introduced a number of different steel systems
of three principal types: close-spaced frame, open-spaced frame, and
load-bearing panels. The Germans did not build as many steel houses
as the British, but they evolved more systems. Their development
of concrete construction involved the introduction of various aggre
gates such as clinker, foamed slag, and pumice, which were often pre
cast into large story-height wall panels to be hoisted by crane into
place within a structural steel framework. The way in which con
crete and steel were used was largely influenced by the fact that
much of the housing built was in the form of apartment houses going
to three or more stories, and thus prefabrication was put on a much
wider basis than in Britain, where it was restricted largely to single-
family houses, or in America, where the concentration was on single-
story, single-family houses. British writers conclude that "when pre
fabrication is thus applied to large buildings it escapes the stigma of
cheapness and nastiness and the development is of much more funda
mental importance, invading as it does the whole fabric of the build
ing trade." 25 It should also be mentioned that by 1929 Hugo Stinnes
and Hugo Junkers, two major German industrialists, were consider
ing industrial methods for improving housing production. Junkers,
the expert in aeronautics, was making elaborate experiments in air
plane-like stressed skin construction.

Unfortunately the fine pioneer work of the Germans with metals
and concrete, much of which was leading to substantial cost reduc-

25 Harrison et al., op. cit., p. 8.

tions, was curtailed by the depression in 1932 and completely halted
when the Nazis came to power in 1933. The use of steel for house
building was prohibited, and all civil building was limited by the
channeling of resources to the construction of fortifications and mili
tary and party buildings. In addition, the rational approach to build
ing was condemned, and it was decreed that "Germanic" ideals were
to be reflected in new construction.

3. France

The French, faced, like the British, with a well-rooted conventional
architecture, appear to have begun their prefabrication efforts rela
tively late. Their structural engineers were making great advances,
but in such other directions as the development of reinforced concrete.
In the year 1927, however, the steel industry started to sponsor sys
tems which used steel for interior and exterior wall surfaces. Some
of the units were similar to those introduced in America in the early
thirties, and there were also steel frame structures and structures of
stucco on metal lath, the last being developed with more success
than marked the British projects of the same type. A development re
lated to prefabrication was the move to standardize and coordinate
the dimensions of various building elements, which received an early
start in France, for beginning in 1929 the Ministry of Commerce pro
moted a campaign towards this end. This was of importance to pre
fabrication in its broadest sense because it was pointed towards the
elimination of cutting and fitting at the site and a consequent relative
shift of work to the factory. Several architectural competitions in
the thirties, out of which emerged a number of proposed prefabrica
tion systems, revealed that the ideas of standardization and modular
design26 were gaining acceptance to a substantial degree.27

2« See below, p. 24.

27 Probably the most widely known of the French prefabrication efforts was
the system of the engineer Eugene Mopin, used in the construction of multistory
apartments in several large projects in and about 1934. The Mopin system, a
combination of prefabrication and site fabrication, consisted of a light steel frame
encased in concrete with intermediate posts of precast reinforced concrete and
external walls of precast vibrated concrete slabs keyed into the posts. Though
the buildings have been criticized for poor sound- and thermal-insulation quali
ties, they were perhaps the most significant experiment of the decade in precast
construction.

4. Sweden

Meanwhile, in Sweden designers made use of local materials to
meet their problems. In contrast to Britain and France, Sweden had
an abundance of timber and prefabrication systems were evolved in
terms of this material and, in the early years, a not very economical
use of it. By 1923 technicians had introduced prefabricated houses
of wood to meet the extensive housing shortage which developed
after World War I. A major interest in the Swedish prefabrication
experience lies in the role of the municipal government of Stockholm
which, through its planning of land use, its provision of credit, and
its self-help plan, encouraged the use of prefabrication for the hous
ing of families of low and moderate incomes. The most interesting
feature of this program was that the occupying family itself often
supplied the unskilled labor needed in the building process. The city
provided plans and skilled labor, such as carpenters, plumbers, and
electricians, for a reasonable fee and furnished guidance to the
family, who made the excavation, laid up the cement-block basement
walls, helped the skilled craftsmen erect the shell and install the
utilities, and carried out much of the final finishing work.

The prefabrication systems were all similar in their main character
istics: wood framed, load-bearing panels surfaced externally and in
ternally with vertical tongue and groove boards and filled with an in
sulating material such as sawdust. This was hardly a pattern suited
to lands where wood is a relatively scarce or costly material (and
this includes even the United States). The panels were delivered
complete with doors and windows, were the full height of the house,
and came in various widths to suit a number of designs. They were
in general extremely heavy. Standards were high, and maintenance
has since proved economical. Involving as it did municipal owner
ship of land, municipal home financing, and municipal provision of
many building services, this program represented the most compre
hensive public assistance to prefabrication to that date.

The self-help scheme was successful enough to persuade private
contractors to offer the same service to homebuilders who were plan
ning to live on either privately or municipally owned land; "in fact,
the majority of small houses built in the garden suburbs on the 'self-
help' plan have been constructed by private builders. But having
pioneered in this method, the city continues its program, constructing
on an average of three hundred dwellings a year/' 28

28 John Graham, Jr., Housing in Scandinavia (Chapel Hill: University of North
Carolina Press, 1940), p. 67.

More than 3,000 self-help houses were built under the municipal
plan alone between 1927 and 1940.29

B. Experimentation and Small-Scale Development in America

Returning to a consideration of what was happening in America
during the twenties, we note at once the difference in the role of
government as compared with Europe. Whereas abroad there were
various types of public stimulus to prefabrication— public housing,
government-supported research and development, government en
couragement of modular design— here there was, except for a small
simplification and standardization program of the U. S. Bureau of
Standards, no federal interest in prefabrication as such, or in re
lated developments. We were in the midst of prosperity and a record-
making building boom. Neither the government nor any of the big
corporations associated with the building industry had reason to push
prefabrication, and consequently development in this field was car
ried on by a handful of crusading individuals and small companies
with limited financial resources.

1. Work in Concrete

In the first part of the decade the interest was primarily in the ap
plication of concrete to small-house construction both by on-site meth
ods—which really involved little or no prefabrication— and by use of
precast elements (see Figure 2). Instances of the latter type were
several systems of story-height units of precast reinforced concrete:
Armostone (1920), Moore Unit (1920), and Tee-Stone (1923). Only
the last of these systems involved more than the walls of the dwelling;
floors, ceilings, and roofs were handled in a conventional manner, and,
as was the case with almost all the early proposals, no attention was
devoted to the mechanical equipment of the house. It is probable
that a total of not more than 500 houses was built by means of these
systems. Meanwhile Grosvenor Atterbury was continuing his work,
which from 1919 to 1921 was being carried out in a laboratory sup
ported by the American Car and Foundry Co. The conclusion

*'Ibid.,p. 59.

reached here was that while a casting process was a sound solution
to the problem, further development by a non-profit agency was
necessary before a commercial enterprise could be successfully under
taken.

It should be noted that while these and subsequent developments
in the use of large precast units were proceeding, the role of concrete
in low-cost single-family dwellings increased significantly through the
use of 16" X 8" X 8" concrete blocks. Such blocks, involving but
slightly more prefabrication than bricks, have, through the years,
been accepted to the point where they form the basis of more than
one-tenth of our annual housing production.30

2. Research by Bemis

It was early in the decade (1921) that Albert Farwell Bemis, a
Boston industrialist, began the sponsorship of research in prefabrica
tion. Through Bemis Industries, Inc., Mr. Bemis owned and con
trolled a number of concerns manufacturing building materials and
products. Among these, the Housing Company was equipped to
fabricate and erect houses and other buildings by either conventional
or novel means, while Bemis Industries itself maintained a laboratory
and staff devoted to research in housing. For the next 10 years, a
period during which prefabrication was quite removed from the
limelight, Bemis Industries, Inc., studied building materials and struc
tural methods in its laboratories and in the field, experimenting with
a large number of different types of construction.31 Its research pro
gram comprised three stages: development of a scheme on paper;
laboratory construction and testing of a full-size section; and finally,
if justified, the building of a house to test the new method for physi
cal performance and cost. Although the program proceeded by fun
damentally logical considerations from one scheme to the next, it is
perhaps fair to remark that the successive attempts were too little re
lated to one another. The lack of continuity in approach may be
noted when we consider that the 22 systems which were tried in
cluded such elements as solid wood panels, plywood panels, concrete
poured in situ, precast gypsum blocks, precast gypsum slabs, gypsum
tubes, an excelsior-magnesite material known as "Acoustex," steel

30 Bruce and Sandbank, op. cit., p. 40.

31 For a summary of this work, see John Burchard II, "Research Findings of
Bemis Industries, Inc.," Architectural Record, 75 (January 1934), 3-8.

PIE-CAST SLA5S

TIE I

EXTUIOH SU&

&USPACC

Figure 2. Precast Concrete Systems of the 1920's:
(1) Frank Lloyd Wright

MAILING SIMPS
ATTACHED WHEN
UNIT IS CAST

PfcE-CAST

fcEIH FOUCED

TEE UNITS

Figure 2. Precast Concrete Systems of the 1920's:
(2) Tee-stone

frames, and steel panels— a pretty fair sampling of all the then-known
construction materials. During the twenties, so far as we know, Bemis
Industries, Inc., spent more time, money, and effort on this type of
research than any other single organization. With the advent of the
depression, however, it was forced to curtail its activities somewhat,
and from 1931 until Bemis' death in 1936 effort was concentrated on
developing new materials in the laboratory (particularly a material
which would at once provide structural strength, insulation, and sur
face finish) and on the development of his cubical modular method
of design. If none of the systems developed by Bemis Industries,
Inc., was ever exploited commercially on a large scale, it is nonetheless
true that its contribution was a significant one, for the development
work on materials and structural methods, particularly on joints,
provided a good deal of practical material for those who were en
gaged in technical problems.

Mr. Bemis' cubical modular method of design evolved from his
work towards better and more flexible coordination of structural com
ponents. He concluded early in his researches that a fundamental,
all-inclusive basis must be established which would coordinate the
dimensions of all structural components, building materials, and in
stalled equipment. The cubical modular method was developed as a
theory of design, but simultaneously its practicability was proved by
applying it to a variety of materials and constructions in experimental
houses which were built and sold. This objective explains some of the
discontinuity in construction ideas referred to in the previous para
graph.

A theoretical discussion of the cubical modular method is given in
The Evolving House,32 Vol. Ill, Rational Design. The method re
quires that the space occupied by the building be considered as a
continuum of cubes formed by parallel lines in each of the three di
mensions, and spaced on a standard module for building layout and
assembly details. Mr. Bemis showed that, in order to permit a maxi
mum of freedom, the basic module should have a length of the order
of magnitude of a wall thickness, and he chose 4" as the unit most
consistent with existing products and practices (as, for example, the
16" spacing of studs in wood frame walls ) . He further demonstrated
that the 4" module could provide the basis for a sound standardiza
tion of all dimensioned building products with at least as great flexi
bility of building layout as was available with former "stock" sizes.

32 Albert Fanvell Bemis and John Burchard, 2nd, The Evolving House: Vol. I,
A History of the Home; Vol. II, The Economics of Shelter; Vol. III. Rational
Design (Cambridge: The Technology Press, 1933; 1934; 1936).

The first large commercial application of the method was made in
1937 by Homasote Co. with technical assistance from Bemis Indus
tries, Inc. The modular details developed for its Precision-Built con
struction enabled Homasote to produce any house designed on the
4" modular basis by relatively simple jig cutting and assembly methods.

In 1938 the heirs of Mr. Bemis founded Modular Service Associa
tion as a non-profit corporation to help the building industry in de
veloping dimensional coordination. The industry effort was organized
under the voluntary committee procedure of the American Standards
Association and is known as ASA Project A62, with the American In
stitute of Architects and The Producers' Council, Inc., as joint sponsors.
Through Project A62 the industry has adopted the cubical modular
method as the American Standard Basis for Dimensional Coordina
tion, with appropriate changes in terminology. The method is called
"modular coordination/' and the "modular lines used as a design
matrix" have become the "standard grid to which building plans and
assembly details are referenced." The standard module is 4". The
objectives of modular coordination are discussed on p. 81.

Through his work on modular design Bemis gave impetus to a
much-needed movement in building, one which was to serve prefabri-
cation through the elimination of much cutting and fitting at the site
and which was to find added support from architects, building mate
rials manufacturers, and the government as time went on. Last, but
not least, was his contribution in The Evolving, House, an exhaustive
study which, in treating modular design and prefabrication seriously
for perhaps the first time, gave these ideas real form and stature.

3. Early Steel Systems

During the latter part of the decade several steel frame systems
were introduced, but these were not of great significance. Those
sponsored by the McClintic-Marshall Corporation and the Gary Struc
tural Steel Corp., for instance, entailed prefabrication only of the
framing members and used these with the close spacing typical of a
wood frame structure. Consequently, little if any economy was
achieved; instead the emphasis of the proponents of such systems
was on the superiority of steel over wood from the points of view of
strength, fire resistance, and dimensional stability. Furthermore,
since the interest of the sponsor did not usually extend beyond the

frame, there was the problem of overcoming the inertia of builders
and of persuading them to depart from established practice for only a
part of the structure, especially when the use of collateral materials
may have offered some problems. The steel framed house of more
recent design tended to use this material economically, taking ad
vantage of its various properties and more efficiently integrating the
frame with the rest of the structure.

4. The Radical Approach

One of the most interesting designs of the period was Buckminster
Fuller's Dymaxion house (1927) (see Figure 8)— interesting not so
much because of the details of the house itself, which in its original
form never progressed beyond the model stage, as because of the ap
proach to the problem. Fuller serves as the symbol of a group of men
who have thought of prefabrication in quite basic terms and have
emerged with the conclusion that the design of the house must be
fundamentally altered if we are adequately to meet the housing prob
lems of our civilization— that, in certain respects at least, revolution
rather than evolution is necessary. Such a group should include,
among others, those who have speculated about houses suspended
from a central mast: the Bowman brothers, George Fred Keck, Eero
Saarinen, Richard Neutra, Peter Pfisterer, and, of course, Fuller him
self; about externally suspended houses: Paul Nelson, Keck, and Leland
Atwood; about hemispherical houses of monocoque construction:
Martin Wagner and Wallace Neff; about various types of mobility in
housing— the trailer house, the folding house, the sectional house: Cor-
win Willson, William B. Stout, Temple H. Buell, Carroll A. Towne,
Carl Koch, and John Bemis. This is not a complete list, nor were all
these designers thinking in terms of low-cost housing or the indus
trialized production of housing, although most of their schemes did
involve a good deal of prefabrication. The kinship they bear to
Fuller is in their attitude towards design, and it is this attitude that is
the important thing about the Dymaxion house, not that it was to be
suspended by wires from a central mast, or that it was to be hexa
gonal in plan, or that it was to be air conditioned, have an automatic
laundry, and a self-contained waste-disposal unit. Indeed, it was as
an attitude that Fuller himself later characterized the house:

An attitude to think truthfully. To think truthfully in the terms of the
latest achievements of the intellect, quite unfettered by history's relatively

temporary national, political and aesthetic bonds. Such bonds are not
habits of thinking but habits of not thinking.38

Looking back,34 Fuller has explained that behind the design of
Dymaxion I lay an effort to maximize the performance of the house
per pound of material in its structure. This objective led to a search
for the means of enclosing the maximum volume with the minimum
surface, for ways to use light materials, and for a structure which
would utilize metals in tension rather than compression in order to
take greatest advantage of their strength properties. Some have held
that Fuller was not as rational as he supposed. Lewis Mumford
pointed out, for instance, that "though Mr. Fuller . . . believes that
he has swept aside all traditional tags in dealing with the house, and
has faced its design with inexorable rigor, he has kept, with charming
unconsciousness, the most traditional and sentimental tag of all,
namely, the free-standing individual house. If we are thorough
enough in our thinking to throw that prejudice aside, too, we may, I
suspect, still find a place for the architect in modern civilization." 35
Another aspect of Fuller's thinking that has been questioned is his
pronouncement of performance per pound as a figure of merit for
house design. Why, it has been asked, per pound of house? What
if it should cost more to produce and use the light metals Fuller calls
for than to fabricate and transport somewhat heavier materials? How
important is transportation cost in the final cost, and to what extent
does transportation cost depend upon bulk rather than weight? Per
haps the reason that the designers of houses have not thought in
terms of performance per pound is that they are not so deeply con
cerned with gravity as are aircraft and ship designers. But this is
not the place to examine the validity of details of Fuller's argument.
The important thing is that he should have thought in terms of some
figure of merit and in terms of what technology had provided and
could provide in materials and structural methods.

At the time his house was introduced, Fuller writes,36 he extrapo
lated curves of industrial progress, of housing demand and supply, of
invention gestation, of the range and frequency of per capita travel,
and concluded that the house, with all the improvements in technol
ogy that would take place in the meantime, could not be industrially

33 Buckminster Fuller, "Dymaxion Houses: an Attitude," Architectural Record,
75 (January 1934), 10.

34 R. Buckminster Fuller, Designing a New Industry: A Composite of a Series
of Talks (Wichita, Kan.: Fuller Research Institute, 1946).

35 Lewis Mumford, City Development: Studies in Renewal and Disintegration
(New York: Harcourt, Brace & Co., 1945), p. 73.

36 Fuller, Designing a New Industry, p. 24.

produced for some 21 or 22 years, until 1948-1949. For a while, in
the last few years, it looked as though Fuller's prognostication might
have been startlingly accurate. A new version of the Dymaxion house
was prominent in the news. Basically the same as the 1927 design,
it had been made round instead of hexagonal and had been lowered
on its mast and fitted with a ventilator on top; recent developments
in light metals, in synthetics, and in aircraft production techniques
were to be applied to its manufacture; it was even accorded a "better
than even chance of upsetting building industry." 37 But this later
Fuller house never got into production, and changes introduced after
Fuller left the company did not help the situation.

It is clear that in 1927, even if the technology had been capable of
it, no one was in a hurry to produce anything so revolutionary. The
building industry had just finished one of the biggest years in its
history and had already passed the turning point of its boom. There
was little talk of a housing shortage; in fact it seemed to industry that
plenty of houses, if not too many, were being produced, even if they
were not going to those who needed them most. Although the ideas
of Fuller and other members of this group of radical thinkers were not
realized in production, they still served two ends: they caused the
architects and engineers to think more deeply about house design,
and, perhaps not so happily, they caused considerable public excite
ment. The outburst of inventions and publicity really arrived, how
ever, with the thirties, when the nation, struggling through a depres
sion, turned anxious eyes towards the technical world in the hope
that some mass-production miracle might occur.

IV. 1930-1940: Prefcibrication Attains the Status of a Move
ment

A. The Background Influences

It was in the early thirties that prefabrication became a widely
recognized movement, and interest in one aspect or another of the

37 "Fuller's House," Fortune, XXXIII (April 1946), 167. But see also:
"What became of the Fuller house," Fortune, XXXVII (May 1948), 168.

idea spread to a much wider group than the handful of inventors and
small companies which had previously been concerned. The spread
of this idea may be attributed to a confluence of factors, economic,
social, and technical.

1. Economic Factors

There was, first of all, the overwhelming effect of the depression,
the impact of which stimulated the search for new kinds of employ
ment and investment opportunity. Though builders and mortgage
institutions were not yet concerned about a housing shortage, it was
clear to many who were casting about for new markets that a radically
low-cost house would offer just such an opportunity. It was generally
recognized that the purchase of a new house was beyond the means
of at least half of the families in America.38 Here was a market if only
one could provide the product. But not only was there a search for
new investment opportunities; it was also necessary to find an out
let for the potential output of existing investments in plant and equip
ment. The steel industry, for instance, operating at one-quarter of
capacity,39 looked desperately for a new market to absorb what it
was capable of producing. Similarly, some of the large building ma
terials producers sought to get housebuilding, which had slumped to
10% of its 1925 peak,40 out of the doldrums. The consequence was a
widespread development of themes similar to that of an article in Col
liers entitled, <cWe Can Build Our Way Out/'41 which called for a
new house manufacturing industry to end the depression.

38 In the mid-thirties, when a mass of statistical investigation began to pro
vide us with some disturbing facts about our economy, it turned out that, if
one took the crude rule of thumb that a home buyer's income should equal half
the cost of his house, some 79% of American families could not afford a "low-
cost" house priced with lot at $4,000 (Family Expenditures in the United States,
National Resources Planning Board [Washington, 1941], Table 1, p. 1).

39 1936 Supplement, Survey of Current Business, U. S. Department of Com
merce, p. 118. In 1932 ingot production was at 20% of capacity; sheet steel
production at 25% of capacity.

40 Total non-farm dwelling units started in 1925: 937,000; in 1933: 93,000
(Housing Statistics Handbook, Housing and Home Finance Agency [Washington,
1948], p. 2).

"Collier's, 91 (June 10, 1933), 12 ff.

2. Social Factors

Added to this general economic outlook was a social atmosphere
in which there were, on the one hand, those whose faith in our eco
nomic system had been considerably shaken and who argued that at
least in housing the government must take an active role in provid
ing for the lower-income groups, and, on the other, those who were
convinced of the basic soundness of a private enterprise economy.
The industrialization of housing was a challenge to those who be
lieved in the private industrial system and thought it could be made
to work in all areas for the benefit of all the people. There were
other relevant social trends: the increasing concern about economic
insecurity, the movement of employment opportunities, and the mo
bility of the population; the increasing scope of government activities;
the growth of a housing movement. These will be discussed more
fully later.

3. Technical Factors

There were also technical developments during this period which
deserve brief mention here: the progress in plywood manufacture
brought about by improvements in glues and veneer cutting; the
better utilization of wood seconds and wood waste to make plastics,
wallboards, and hardboards; the expanded production of other sheet
materials made from gypsum, asbestos, cane fiber, newspaper, etc.;
the development of sheet steel and the continuous strip and cold-
rolled processes; the improvement in alloys, especially of the light
metals; and the treatment of cementitious materials by vibrating,
aerating, and use of lightweight aggregates. Again, these will be dis
cussed later.

This complex of economic, social, and technical factors will be
analyzed here by summarizing the activities of those groups which
concerned themselves with broad applications: the non-commercial
research organizations, the government, and the business and financial
world.

B. Non-Commercial Research and Development

One aspect that distinguished American research and development
in prefabrication from those in other countries was their continuity.42
In America, as elsewhere, commercial sponsors were active, in gen
eral, only if they saw profit possibilities. But here, unlike most other
countries, there was a core of constant activity in research which was
carried on with the principal object of providing better and more
economic shelter. True, the scale on which development and ex
perimentation were carried out varied with business conditions, but
at least a small amount of effort was consistently expended regardless
of the immediate economic problems at hand.

1. Pierce Foundation

In the twenties there were Bemis and Atterbury, and now, in the
thirties, other organizations entered the field. The Housing Research
Division of the Pierce Foundation,43 in Raritan, N. J., was founded in
1931 and under the direction of Robert L. Davison began a search for
materials and structures that would yield a house of lowest possible
cost consistent with adequate physical standards. Among the mate
rials which this group tried were concrete, plywood, composition
board, cellular glass, stabilized earth, and a hydro-calcium silicate
composition known as "Microporite." Behind much of its experimenta
tion lay the same aim that motivated Bemis: to find a single material
which would serve both as structure and as enclosure. A number of
test houses were erected. The first, in 1932, had an open-spaced
(12%') steel frame and floor-carrying walls of welded lattice trusses
encased in a cementitious material, a system intended primarily for
multistory apartment construction. A second experimental house
(1935) also used an open-spaced steel frame with precast reinforced
Microporite slabs for walls, floors, roof, and partitions. Both these
structures used panels horizontally, a type of design which the Founda
tion tended to favor for its flexibility in planning and fenestration, de-

42 Harrison, op. cit., p. 124.

43 The John B. Pierce Foundation of New York City was chartered in 1924 to
carry on educational, technical, and scientific work in the general fields of
heating, ventilating, and sanitation. It was endowed in the will of John B.
Pierce, New England industrialist and financier.

spite the much more common preference for vertical elements. Later
a community of 20 plywood houses in Highbridge, N. J., was built to
permit a continuous study of family needs and maintenance prob
lems.

The Foundation has also done considerable work on plumbing and
heating equipment and was largely responsible for the integrated
mechanical core as it first reached the market in 1935 in the American
Motohome. Its work was reflected commercially in at least two
other ways. The studies of floor plans and family living habits were
at least partially responsible for the 24' X 28' single-story house
which has to a large extent become "standard" in the low-cost field.
A second instance was the Cemesto House, released for commercial
development with considerable success in war housing.44 This dwel
ling had an open-spaced wood frame clothed with horizontally laid
slabs of Cemesto, a sandwich material composed of a cane fiber insu
lating core faced on both sides with cement asbestos sheets. At the
present time the Foundation is continuing its work on a number of
phases of house construction, concerned particularly with systems of
prefabrication employing stressed skin plywood panels in connection
with structural frames of either light-gauge steel or wood.

2. Universities

Purdue University's Housing Research Project, instituted in 1935,
was another non-commercial agency that carried on work in the field,
much of which was done in cooperation with industry. One of its
early efforts was the building and testing of five commercial types of
low-cost house, two of which were prefabricated. Engineering and
cost studies were made and published. Considerable work has also
been done in the fields of heating and ventilation.

Other universities were, of course, also conducting research in re
lated areas, yet this work was generally not focused specifically on the
problem of building the house shell, but was more often concerned
with various types of economic studies, with family needs, or with
mechanical equipment. This was in part a reflection of the peculiar
organization of the housebuilding industry which left it, by compari
son with other industries in, for instance, the chemical and electrical

44 For instance, the community built for the employees of the Glenn L. Martin
Company near Baltimore, Md.

fields, quite unable to pose the problems, encourage the research, and
utilize the results.

C. Government Activity: Techniques and Standards

1. U. S. Forest Products Laboratory

Several government agencies played a prominent role in the de
velopment of prefabrication during the thirties. One of these was the
U. S. Forest Products Laboratory,45 whose purpose it was to study the
utilization of our forest resources and which had for some time been
working on various types of glue and plywoods. Later it began to
work on house construction, and in 1935 its first stressed skin plywood
house was built, embracing a structural design that was to have a
very great influence on the development of the industry. The stressed
skin principle was not new, except to housebuilding; the idea was
simply to build the wall panel as a box girder and thus use the sur
faces of the panel in such a way that they, as well as the framing
members, would carry a major part of the load. Though not new,
the principle waited for its housing application upon the creation of
the proper plywoods and glues. Stressed skin construction offered
good possibilities for saving material, mechanizing wood fabrication,
and lightening the structure, and it was therefore eagerly adopted by
a number of prefabricators and was extensively exploited in war hous
ing. The contributions of the U. S. Forest Products Laboratory to
prefabrication had really just begun, however, for in the ensuing
years the research carried out there dealt with many of the technical
problems besetting manufacturers, for example, dust patterns, inter-
wall condensation, and the bowing of panels due to changes in mois
ture content; and the work of the Laboratory with new materials and
production techniques had great import for most of the firms in the
industry. It is probably pertinent to remark that this publicly spon
sored research organization served a unique role in an industry char
acterized by small companies which were generally incapable of
carrying on any extensive research of their own.

45 Madison, Wis. Established in 1910, and operated by the U. S. Forest
Service, Department of Agriculture.

2. National Bureau of Standards

Another government agency which rendered technical assistance
to prefabrication was the National Bureau of Standards, which in
1937 began a program of research in building materials and struc
tures for use in low-cost housing. Testing procedures for such ele
ments of the house as walls, partitions, floors, and roof were developed,
and a large number of reports on the physical properties of various
materials and systems of construction, some of them prefabricated to
a large degree, have since been issued.46 Work was also done on
plumbing and heating equipment, on Simplified Practice Recom
mendations,47 Commercial Standards,48 and building codes. Ulti
mately this program of performance tests and related building studies
may have a large effect on the writing of codes and specifications
and on the whole development of better and cheaper methods of
construction.

3. Federal Housing Administration

It is the Federal Housing Administration, however, which has prob
ably been the most important single influence in setting standards
for the construction of low-cost houses. Through its guides for rating
mortgage risk the FHA established many criteria for house construc
tion. Prefabricators who were trying to tap medium- and low-cost
markets with a new product had to rely to a considerable extent on
FHA mortgages for home financing. When a prefabricator's house
was approved on a technical basis, the Washington office of the FHA
issued an Engineering Bulletin proclaiming that fact and giving perti
nent data. One result of this was to create a basis for evaluating the

46 Building Materials and Structures Reports, U. S. Department of Commerce,
National Bureau of Standards (Washington, 1939+ ).

47 Aimed at eliminating waste through the establishment of standards of prac
tice for stock sizes and varieties of specific commodities that are currently in
general production and demand. Adopted voluntarily by industry with the
assistance of the National Bureau of Standards.

48 Aimed at establishing standard methods of test, rating, certification, and
labeling of commodities and at providing uniform bases for fair competition.
Adopted voluntarily by industry with the assistance of the National Bureau of
Standards. Commercial Standard CS 125-45 for Prefabricated Homes, accepted
in 1945, was the first applying to the prefabrication industry. A revision,
CS125-47, was brought out in 1947.

prefabricating systems commercially available. Yet, while the FHA
has had a distinctly salutary effect in revising mortgage financing, in
inducing banks to lend on prefabricated houses, and in establishing
standards, the influence of its approval has grown so great that a new
house manufacturer is now apt to be severely penalized without it.
The importance of this institution increased steadily throughout the
thirties, but it was only in the postwar period that it reached its pres
ent tremendous significance. It is perhaps not too much to say that
FHA approval is now a matter of life or death to the prefabricator
about to enter production with a new system, and in view of the power
wielded by the FHA, its policy with respect to new developments in
building has become a matter of considerable importance.

D. Government Activity: Prefabricated Construction

When we turn to a consideration of the various government agencies
which entered into housebuilding directly, we find vast differences in
approach, deriving chiefly from equally large differences in purpose.
At one extreme there was the public housing program, first under the
Public Works Administration and then under the United States Hous
ing Authority, which was conceived in terms of highly permanent
fireproof multifamily buildings having low maintenance, high physical
standards, and long-term (60-year) amortization to make rents as
low as possible. The program was aimed at rehousing slum dwellers
and usually involved building in dense urban areas. For these and
other reasons the public housing authorities did not regard prefabri-
cation very seriously and did not use it at all until the advent of the
war housing program.

1. Farm Security Administration

In contrast to this public housing program, the Farm Security Ad
ministration 49 was charged with promoting self-sufficiency, decreas
ing tenancy, and resettling migrants on the land. The FSA's ap
proach was from the beginning characterized by a willingness to ex-

49 Formed in 1937 to carry on the work of the Resettlement Administration
and several other farm agencies.

periment, and its technical staff, which was decentralized into 12 re
gional offices and by comparison with the public housing program was
relatively free from Washington control, emerged with many fresh
approaches to construction. In 1938 the FSA built 100 farmsteads for
sharecropper families in Missouri and achieved very low costs by
prefabricating wall and roof sections and making use of a large pro
portion of unskilled labor. The price was $1,105 for a 24' X 36' five-
room house, and $2,000 for a farmstead including house, barn, stor
age shed, privy, fencing, roads, and a well.50 The following year 60
farmstead units of steel were ordered from the Tennessee Coal, Iron
& Railroad Co., a United States Steel Corporation subsidiary, and were
erected in Georgia, Alabama, and South Carolina. The houses, a
little smaller than the wooden ones built in Missouri, cost 503?
more.51 In the west the FSA built whole communities for migratory
farm workers and carried on a number of experiments with plywood
and other unconventional construction in an attempt to provide mi
grants with something more substantial and more economical than a
tent. Some 6,000 steel "minimum" units, 12' X 14', were built for this
purpose. Though not related specifically to prefabrication, the plan
ning and building of these integrated farm communities were among
the agency's most significant contributions in the field of housing.
By the end of 1940 the FSA had built more than 26,000 individual
houses,52 and at that time The Architectural Forum could write, "To
day in face of a national emergency, Farm Security stands out as the
agency most experienced in the work of building houses quickly and
cheaply." 53 Its work continued into the war period and included in
a dormitory project near Vallejo, Calif., one of the first uses of stressed
skin plywood construction in two-story buildings.

2. The Fort Wayne Experiment

Still another approach was the widely publicized Fort Wayne Plan
developed jointly in 1938 by the FHA, the Fort Wayne Housing Au
thority, and the PWA. The 50 single-family units comprising this

*°The Architectural Forum, 69 (November 1938), 393-4.

51 The Architectural Forum, 70 (January 1939), 68; Architectural Record, 85
(January 1939), 38-9.

52 Including those built by the agencies the FSA had taken over ( The Archi
tectural Forum, 74 [January 1941], 13).

™The Architectural Forum, 74 (January 1941), 3.

project were built of stressed skin panels and rented for $2.50 a week.
WPA labor was used in a factory rented and equipped by the Fort
Wayne Housing Authority, and houses were placed in blighted areas
on vacant lots which had been bought for $1.00 each, the seller be
ing given an option to repurchase. This plan differed significantly
from the public housing formula, and it inspired considerable con
troversy.54 Its 480 sq. ft. houses were below USHA standards; it re
quired demountable units in case the land was reclaimed by the
former owner; it called for lighter construction and shorter amortiza
tion periods; it used WPA labor; and it involved private, insured
financing. Perhaps the fact that this plan was not adopted elsewhere
is evidence that it was not suited to the conditions and the times. It
did, however, bring attention to another example of prefabricated
construction.

3. Tennessee Valley Authority

Probably the most important government effort in prefabricated
building during this period, at least from a developmental standpoint,
was the work of the Tennessee Valley Authority with demountable
sectional houses. Actually this work extended well into the war years,
but it had its origin in an earlier period and is therefore discussed
here. The TVA required temporary housing for the construction
workers on its many hydroelectric projects and had for some time
speculated about making good portable houses as an alternative to
building mere shacks.

In 1934 Louis Grandgent, then chief of TVA's architectural section,
proposed a scheme for building a house which could be separated
into four or five sections each of such dimensions that it could travel
safely by truck and trailer over public highways. After some ex
perience with transporting conventional houses by barge, the sectional
house idea was developed by the TVA staff under the supervision of
Carroll A. Towne. In 1940 the first TVA sectional houses were built
for transportation by truck to the site.55 Sections measured 7%' X 22'

54 See, for instance, Charles Abrams, "Fort Wayne Housing Plan Analyzed,"
American City, LIV (April 1939), 106; National Association of Housing Officials
News (February 28, 1939), 9-11; (March 28, 1939), 21.

55 In 1939 General Housing Corporation, Seattle, Wash., had begun manufac
ture of a sectional but otherwise conventional wood frame house. The four-room
dwelling was made in two 12' wide sections, completely finished and equipped

X 9%'» were of wood frame construction, and weighed three tons.
They left the factory with all electric, heating, and plumbing equip
ment installed, and arrived at the site completely finished even down
to light bulbs and screens. Houses were finished at the site in as
little as four hours by bolting together two or more sections. The
next year the TVA's design was adopted by the Federal Works Agency
for war housing, but in order to meet nationally standardized require
ments, a pitched roof, hinged to let down during transit, was added.
In 1942 the TVA began experimenting with designs that frankly
recognized the house section as a trailer and used certain aspects of
trailer construction. Weight was reduced by the adoption of stressed
skin principles, and transportation costs were cut from 30^ per sec
tion per mile to 23^. These houses were trucked as far as 600 miles.
Still later, when the Army erected several thousand of these houses
at the atomic bomb project at Oak Ridge, Tenn., a boom crane was
used to put sections in place instead of the jacks and rails that had
previously been used. This kind of prefabrication, which reduced
site labor to as little as 5-10% of total direct labor, was later to have
considerable influence in several postwar designs, notably the Prenco
and Reliance houses and the British AIROH house.

E. Commercial Development by Private Enterprise

While a number of non-profit and government institutions made
significant contributions to the development of prefabrication, its
widespread adoption on a commercial basis awaited the efforts of
private enterprise, and it remained for some entrepreneur to organize
a successful pattern of operations. In the early thirties one element
of the business world, big business, saw the challenge of prefabrica
tion and talked as though it were prepared to turn the gleam in the
inventor's eye into a profitable operation. For reasons outlined pre
viously, at least a dozen of America's largest corporations entertained
the idea for a while, and a few stuck with it continuously. Among
the big names were United States Steel Corporation, Great Lakes
Steel Corporation, American Car and Foundry Co., Pullman Standard
Car Mfg. Co., The Celotex Corporation, Johns-Manville Corporation,
U. S. Gypsum Co., American Radiator & Standard Sanitary Corp., and

in the factory and trucked to the site (in the Seattle area) where it was bolted
together. Average price, erected and with lot, $3,900; f.o.b., $2,980 (The
Architectural Forum, 70 [April 1939], 286).

General Electric Company. Several of these companies went no
further than to supply certain components to prefabricators in accord
ance with specifications. Some of the materials producers went to
the extent of developing a house system, not infrequently a structure
which reflected the attempt to find every possible use for that com
pany's product. Only a few companies maintained a prefabrication
research establishment that was more than a token effort, and these
few did it on a budget that was meager compared to the research ex
penditures of equally large companies in other industries. However,
most of these companies and a number of others retained advisors to
keep in touch with current developments.

1. General Houses, Inc.

Related to the interest of big business in prefabrication were the
proposals for at least two rather ambitious corporate structures.
The first of these was General Houses, Inc., organized in 1932.
Under the leadership of Howard T. Fisher, General Houses was
to design, coordinate, and assemble standard parts to be produced
for it by a number of prominent specialists. It took its pattern
from the automobile industry, in which the nominal manufacturer
usually acts more as an assembler than as an actual producer but
does assume responsibility for coordinating the many elements in
volved and for providing a complete service to the customer. With
unused capacity available in thousands of plants throughout the
country, General Houses would need no plant and would have
none; instead, parts were to flow from the specialized manufacturers
via warehouses to the site where they would be assembled. The
house was not to be standardized; it was to be custom built from
standard elements. In many respects this pattern was followed,
with the cooperation of several large corporations.56 Research and

56 Among others, Bethlehem Steel Co., Pullman Standard Car Mfg. Co., Curtis
Companies Incorporated, American Radiator & Standard Sanitary Corp., Con
tainer Corporation of America, and Weyerhaeuser Timber Co. General Houses
had an extremely simple corporate structure. Only one or two of the cooperat
ing companies ever held any stock, and their stockholdings were nominal. After
Pullman withdrew to concentrate on the expanding car business, Bethlehem Steel
took over and for several years fabricated all the steel parts used by the com
pany. Most of the millwork and other woodwork was made by Curtis right
through to the war years, when Curtis fabricated complete panels for walls,
floors, ceilings, and partitions for some of General Houses' war housing projects.

development were financed by General Houses itself and standard
parts made to General Houses' specifications were purchased from
suppliers as in the automobile industry. Beginning with a house
of load-bearing steel panels (see Figure 9), General Houses changed
in 1936 to a steel frame system with sandwich panels of cement
asbestos sheet (exterior), insulation, and plywood (interior), largely
in an effort to achieve lower cost. Architecturally the design was
quite modern with a minimum of ornament, a flat roof, smooth
exteriors, vertical battens, and a good deal of glass.57 Up until
shortly before the war General Houses built several hundred of
these houses, an output far below its proposed mass-production
levels. It had encountered a few technical difficulties such as
interwall condensation, none of them serious, but had met with
its largest problems in the realm of financing and marketing. The
dealers came largely from outside the building industry because of
opposition to prefabrication within the industry. In some respects,
they could do a better job of retail merchandising, but they were
severely handicapped by lack of practical building experience. An
other major source of trouble was in obtaining suitable mortgage
appraisals, a problem which derived largely from the unconven
tional nature of the house. General Houses revised its approach
towards the end of the decade, partly for these reasons and partly
because of increasing shortages in steel. Fewer designs were of
fered, wood was employed as a basic material, the roof was peaked,
and the general appearance was made to conform with convention.
About this time the defense housing program started, curtailing
private building to a large extent, and General Houses changed
its plans further, becoming one of the first prefabricators to par
ticipate in the war housing effort and influencing in part the use of
prefabrication in war projects.

2. Houses, Inc.

The second of these grand ventures was Houses, Inc., started
by Foster Gunnison in 1934 at the instigation of Owen D. Young,
then Chairman of the Board of General Electric Company. Houses,
Inc., would build no houses, but it would cooperate with other com
panies in the development of houses of several types. To this end

57 Price in 1934, erected but less freight and cost of lot: four-room house,
$4,500; six-room, two-story house with garage, $8,550.

it proposed to engage in research and provide assistance in the
management and financing of housing enterprises. American Ra
diator & Standard Sanitary Corp. and General Electric Company were
the companies which were to cooperate in the development work.
Houses, Inc., was to own stock in other companies which would
assemble and erect the dwellings. In 1935 there were two such
affiliates, National Houses, Inc., which had a steel-frame steel-panel
system, and American Houses, Inc., which was producing the
Motohome.

The promotion of the Motohome was the biggest activity of
Houses, Inc., that year. Engraved invitations to the first exhibition
announced the "American Motohome" as "the prefabricated house
that comes complete with food in the kitchen." In Wanamaker's
New York store on April 1, 1935, President Roosevelt's mother un
tied the ribbon that bound the Cellophane-wrapped house.58 The
public found a house of steel frame and asbestos cement panels,
with a flat roof, corner windows, and an exterior which frankly ex
pressed the panelized construction. Inside was a mechanical core
including plumbing, heating, and electrical equipment, the product
of work by the Pierce Foundation and General Electric Company.
The promotional campaign included not only store exhibits and
demonstration houses such as those built in Westchester County
that summer, but also extensive publicity in the press; Sunday sup
plements and trade papers alike carried illustrations of what a com
munity of Motohomes would look like and what sort of truck would
carry them from the factory to the site. In late 1935, however, there
were several disputes among the backers 59 and the management of
Houses, Inc., and, for reasons which were primarily personal, Foster
Gunnison sold out his interest and went to New Albany, Ind., to
found his own company. General Electric carried on Houses, Inc.,
for about another year without much success and then liquidated it.

3. American Houses, Inc.

Meanwhile, American Houses, Inc., which had had an independent
existence of its own since its inception in 1933, continued on its

58 500,000 people are said to have visited the Motohome during a six-month
exhibition.

59 By October 1935 General Electric had bought out American Radiator's share
in Houses, Inc., and owned it outright.

way under the leadership of its founder, Robert W. McLaughlin.80
The Motohome, of which only about 150 had been sold, was aban
doned.61 Emphasis was placed on reaching the low-income market
with a conventional product rather than the middle-income market
with a "better than conventional" one.62 In addition, American
Houses radically altered its designs by adopting peaked roofs, wood
sheathing, clapboard exteriors, and plywood interiors. The steel
frame was retained for another two years, until 1938, when it was
discarded in favor of conventional wood framing because the latter
was more flexible, easier to use with other materials, and could be
fabricated with less elaborate equipment in the plant. By the end
of the decade American Houses' system was of traditional platform
frame construction, and the extent of prefabrication had been re
duced to precutting and partial preassembly of panels.63 There was
developing within the company at the same time a trend towards work
ing through contractors who were building projects rather than
selling through dealers to individual customers. While it was thus
changing its pattern of operations, American Houses grew to the
point where, at the beginning of the war, it was one of the leading
prefabricators in the United States.64

4. Gunnison

In the course of the same period, Gunnison had also developed
one of the best-known firms in the industry.65 He had begun in a

60 Now a member of the Board of Directors.

61 Possible reasons: sales resistance to its appearance; its mechanical core re
quired servicing by American Houses; it was overloaded with equipment manu
factured by its original sponsors.

62 The minimum Motohome was a single-story four-room house with garage
selling at $4,950, erected but less lot. Other models were priced up to
$15,000.

<*The Architectural Forum, 73 (July 1940), 69 ff.

Later (1943) American Houses began to describe its business as a "refining
operation," a stage between the manufacturer of raw materials and the con
tractor.

64 One of American Houses' most successful projects was a 136-house sub
division for Bethlehem Steel Co. near Baltimore, Md., 1939-1940. The four-
room house was priced, with lot, at $2,750. At about the same time American
Houses was also building in another price range: garden apartments in New
Rochelle, N. Y., to rent at $20 a room (loc. ctf.).

65 Starting in 1935 as Gunnison Magic Homes, Inc., the name was changed
in 1937 to Gunnison Housing Corporation, and in 1944 to Gunnison Homes, Inc.

modest way, renting a small plant in which to produce his stressed
skin plywood panel houses. The first of these, even though of tradi
tional appearance, aroused a good deal of protest, but this stemmed
mainly from local building people who saw their interests threatened.
Opposition diminished and public acceptance grew as Gunnison and
his houses became known. The 1937 spring flood of the Ohio River
accidentally proved beneficial by showing that the houses, some
of which had been immersed, were of sound construction and by
giving Gunnison an opportunity to compete with conventional
builders in a relief project undertaken by the New Albany Housing
Authority, a project which measurably added to his prestige. Em
phasis had, from the beginning, been on low-cost homes 66 sold
through dealers to the ultimate consumer. There was a brief at
tempt at marketing through the building of housing projects under
the Gunnison Village Plan,67 but the overall trend was in the oppo
site direction, towards the evolution of a system of retail merchandis
ing that would diversify sales risk by making many small sales to
individual customers. Gunnison brought to prefabrication the abil
ity and approach of an organizer and salesman. He was more
determined than most prefabricators have been to break completely
with the traditional operations of housebuilding and to draw his
personnel and his manufacturing and marketing methods from fields
characterized by true mass production. In retrospect his influence
appears as a major factor in shaping one of the important market
ing patterns of the industry.

5. The Nature of Efforts by Big Business

A brief account of the early phases of several enterprises can
hardly do justice to the effort expended on prefabrication in the
thirties, but a few examples may suffice to show the kind and extent
of activity by big and little business.

Regarding the large corporations, it might be said that their re-

66 In 1936 a four-room, 24' X 32' house sold for $2,650 erected but less lot.
The "Miracle Home" offered in 1939 was priced at $2,950; $350 down, $21 a
month.

67 Under this plan a local corporation would acquire land, erect a community of
Gunnison homes, and manage it after completion. 80% or in some cases 90% of
the total value of the completed project was to be provided by an FHA-insured
mortgage. The balance would be represented by a stock issue of which
Gunnison Housing Corporation would own a portion, the rest being held by
local investors.

search was generally biased by some motive other than the indus
trialization of housing; often they designed a house to use the
maximum amount of whatever material they produced. This is not
to deny that the materials and equipment producers, who, of all
the elements in the building industry, were the only ones capable of
financing research on a large scale, did make great contributions in
improving their own products. There also was significant progress
in mechanical, electrical, and heating equipment, and in insulation,
wallboards, flooring, roofing, and glass. But effort at the integration
of parts, research in house design, and studies in construction tech
niques—these were largely neglected. Development work by large
corporations in prefabrication remained embryonic; rarely, if ever,
was an idea carried through to the pilot plant level. On the mat
ter of distribution there was little significant activity. The idea
that an organization for mass distribution of houses would have to
be established before there could be any mass production had nod
ding approval in theory but little application in practice. Grandiose
promotion schemes there were, but these should not be confused with
serious attempts at establishing a marketing pattern, arranging for
financing, meeting code difficulties, and overcoming problems at
the site. This is not meant as an indictment of all the large cor
porations; some put a good deal of effort into developing their
products and struggled for a while with the idea of selling a house,
but ultimately they retired for the most part to selling prefabricated
components, usually for non-residential construction. The sub
sidiaries of Great Lakes Steel Corporation 68 and American Rolling
Mills Company 69 may serve as illustrations.

Yet it is probably fair to remark that the large corporations never
really threw their full resources into the fight. To a certain extent
this may have been a manifestation of the inertia of bigness. But
probably more than this it was a reflection of the organization of the
building industry. There is, first of all, a basic schism of long stand
ing in this industry.70 On the one hand are the manufacturing and
construction interests whose profits stem from the production of
buildings, and, on the other, the rentier and realty interests whose
profits stem from the ownership of buildings (and land). Where a
high rate of production might lead to a disturbance of established
values a conflict of interests is apt to occur, and the position of the

68 Stran-Steel.

69 Steel Buildings, Inc., and The Insulated Steel Construction Co.

70 James Marston Fitch, American Building; The Forces That Shape It (Bos
ton: Houghton Mifflin, 1948), pp. 334-9.

rentier interests who control land and home financing is in this
case a very strong one. Furthermore, the nature of the housebuild
ing process up to now has been such that no single person, organi
zation, or even industry has a sufficient stake in the completed
house (in terms of dollar value) to justify research and develop
ment in the fabrication, assembly, and overall construction of the
house itself and the creation of coordinated marketing processes.71
Because his own stake in the house is small, no supplier makes a
serious effort to reduce the use of his product or service— though
this might lead to greater overall efficiency. Each attempts rather
to increase its use, firm in the knowledge that what he does makes
little difference in the total cost. This attitude has been seen through
out the field, in labor organizations, materials producers, and fixture
and equipment manufacturers alike. There has been no element
in the housebuilding industry with sufficient motivation and with
sufficient power and means of control to initiate fundamental changes
in the fabrication and construction processes and carry them through
to the final product. Another factor in the housing field which might
deter a large corporation from directing its research efforts in that
direction is the likelihood that marketing outlets for the house
package would be in direct conflict with existing marketing outlets
upon which the company might be in large measure dependent.
There was rarely sufficient likelihood of profit to warrant taking such
a risk. In addition, experienced salesmen were well aware that pub
lic attitudes about the home are more strongly entrenched than
attitudes towards other less historic and less emotion-loaded products.

6. The Role of Small Firms

Because of the relative inactivity of the large corporations the
role of the small entrepreneur, that is, the innovator of relatively

71 This point has been made many times, notably recently by C. F. Rassweiler,
Vice-President for Research and Development, Johns-Manville Corporation, in a
talk before the Annual Fall Meeting of The Producers' Council, Inc., New York,
September 30, 1948; and by Robert W. McLaughlin, architect, in a talk at the
Massachusetts Institute of Technology, February 26, 1948. McLaughlin stated
that his studies of a "typical" 750 sq. ft. low-cost house showed that the largest
single element of cost, dimension lumber, represented less than 10% of the
total final cost. By the same token, one may find here one of the keys to a
successful approach to prefabrication— capital aggregations and competent staffs
large enough to bring vertical integration to the whole process of housebuilding.

fest financial resources, was a quite important one. There were
Jy scores in this category, but the vast majority never got be
yond the initial stage of invention. Throughout the decade there
continued the almost naive belief that the invention of some joint or
wall section was the answer to the problem. Some went further
and designed floor, roof, and partition details; a still smaller group
went on to consider the mechanical equipment in the house; and
only a few attempted to outline and organize a pattern of opera
tions that included all phases of the enterprise including distribu
tion. The number of would-be innovators in this period was so
formidable that it would be hopeless to attempt here a review of
even the better half. Instead we shall look briefly at the field as a
whole and point out the broad trends.

F. General Trends and Characteristics

1. Ideas and the Public Mind

One clue to the kind of thinking that was going on may be found
in the Symposium on Prefabrication sponsored in 1935 by Richard
son Wright, editor of House and Garden. Gathered for dinner and
debate were some dozen people who had achieved a certain emi
nence in the field. The discussion ranged over many aspects of the
prefabricated house: its advantage over conventional houses, its
optimum useful life, the problem of financing, the question of modern
design, prefinishing versus site finishing of panels, etc. The steno
graphic report of the Symposium 72 contains some remarks that are
interesting in retrospect:

John Ely Burchard, vice-president of Bemis Industries, Inc.,
Boston:

It is very important that provision be made for financing the houses, and
the prefabricator cannot dump the problem in the lap of the banks. Un
questionably the prefabricated house will be a sounder and more uniform
security but the industry itself must make some arrangements for financing.

Raymond V. Parsons, consulting engineer, Johns- M an ville Corpora
tion, New York City:

72 House and Garden, LXVIII (December 1935), 65-72.

It can almost be taken for granted that when good prefabricated houses
become a fact their architectural style will be different from the quaint
English cottages and Cape Cod Colonials that are the present favorites of
the speculative builders. The idea that we should take new and better
building materials and mould them into the lines and textures of old ma
terials possessing any number of shortcomings is abhorrent.

Howard T. Fisher, president of General Houses, Inc., Chicago:

The final decision, in the matter of design, will of course depend on what
the public wants. But in everything else the public has shown its prefer
ence for the best in modern design, and I doubt if they will pay extra for
faked imitations of the past when they buy their houses. As a matter of
fact, I believe the greatest selling point these houses will have in the next
decade will be their style.

Robert L. Davison, Director of Housing Research, John B. Pierce
Foundation, Raritan, N. J.:

I can't agree with that [Fisher's statement that a 'sloping roof becomes
economically unsound because it is too inflexible']; it depends on the
material used. We have worked with one material where a flat roof
was the only logical solution, and now we are working with a material
which cries out for a pitched roof.

Fisher (on what the useful life of the prefabricated house should
be):

I think if the cost could be correspondingly reduced it should be as short
as possible— up to a certain point! ... It would obviously be more eco
nomical, due to the obsolescence factor, to buy a house that would last, say,
fifteen years, and which would cost only 60% of a house that would last
thirty years, if that were possible. We have been building— this is par
ticularly true of England— for too great a period of time. . . . What
would be the sense in building a refrigerator to last 100 years when you
know improvements will be constantly coming into the market?

This may have been the first symposium on the subject; it was
by no means the last. The press, lay and technical, was very gener
ous in its attention to prefabrication and produced words much
faster than prefabricators did houses. The inventor's desire for
publicity was of course abetted by the editors and apparently even
by the general public, whose curiosity and interest in the home
made it receptive to most of what was said. If a house suspended
from a mast was no longer a sensation, then perhaps a mobile house
was. The influence of the trailer craze that hit America about 1937
spread into housing circles, and it was not long before someone
had figured out how to cure our social ills with mobile dwelling

units. There were houses of copper and of cotton; houses could be
hauled down Main Street or floated down a river; and a hundred
names, from "prefabs" to "motorized zipper housing/' were bestowed
upon these proposals. Probably all this publicity did more harm
than good. It led many people to believe that some miracle would
solve the problem, and at the same time it confused them about the
nature of that miracle, what the prefabricated house looked like,
and where it could be bought and for how much.

2. External Obstacles

Prefabricators encountered a number of external obstacles as they
tried to bring enterprises from the experimental stage into commer
cial production. Such obstacles stemmed in part from local mate-
rials dealers who might decline to sell to the prefabricator's dealer
certain items needed to finish the house, or to grant him a line of
credit; or who might bring pressure on the local building inspector
not to grant a building permit. They were due in part, also, to
organized labor, not so much because of actual fights, although these,
too, occurred, as because of passive resistance and refusal to handle
prefabricated material. The banks and local FHA offices presented
another type of obstacle in their reluctance to lend or insure loans
on prefabricated houses and in their tendency to make low ap
praisals for mortgage purposes. Not the least of a prefabricator's
troubles were those arising from local building codes written in
terms of specification rather than performance, and very often ex
cluding his type of construction. Where the code contained a clause
permitting new types of construction if the building inspector was
satisfied, as the result of a test, that the system was adequate, such
tests had to be made at the prefabricator's expense. Sometimes
one inspector would refuse to accept results of a test made in an
other locality, or there would be a lengthy court case involving a
code issue. Not infrequently the mere prospect of such obstacles
was enough to dissuade the prefabricator from marketing in a cer
tain area, and those firms hardy enough to pioneer in this respect
were required to spend a major part of their energy simply in over
coming the various types of external resistance.

3. Trends in Design

At the Chicago World's Fair of 1933 the modern house exhibit
contained only three dwellings which were prefabricated to any im
portant extent: General Houses' house of steel; Stran-Steel's house
embracing that company's close-spaced frame system, and the
Rostone Corporation's house of precast synthetic stone. This was
not a complete representation of the embryonic house manufactur
ing industry, nor was it especially successful in selling the idea of
prefabrication. At about the same time the files of Bemis Industries
and of United States Steel Corporation's prefabrication advisor con
tained some 40 or 50 American systems which had been proposed
and not yet abandoned. By the end of 1935 The Architectural
Forum, already watching prefabrication as a parent might watch
a precocious child, could list some 33 systems which were supposed
to be commercially available.73 Of these, 16 were steel frame con
struction using panels of various materials such as asbestos cement,
precast concrete, steel, or composition board; five were of steel load-
bearing panels; eight were of precast concrete; one was of precast
gypsum; two were of wood frame; and only one was of plywood.
Of 25 commercially available systems which the Forum 74 reported
in 1938, 15 used steel, two used plywood, and precast concrete was
still a challenge not to be abandoned, with five systems in use.

By the end of the decade a swing away from steel was visible.
On the technical front, the combination of insulation, condensation,
and corrosion problems had at least temporarily defeated many of
the proponents of the steel house. Another and more formidable
obstacle was the problem of selling houses fast enough to justify
a large investment in plant and equipment. No matter how well
steel might be suited to industrialized production methods, these
methods usually required a very substantial capital investment in
manufacturing facilities. Consequently, production at low volume
would be high-cost production, and costs could not be brought down
through mass production until a system of mass distribution had been
established.

Where was the vicious circle to be broken? Perhaps by using a
material that would be economical even at low volume, although it
might lend itself less to industrialized methods and ultimate cost

73 J. Andre Fouilhoux, "Prefabricated Units for the Home," The Architectural
Forum, LXIII (December 1935), 544-76.

""The Architectural Forum, 68 (February 1938), 66, 70.

reduction. Wood was such a material, and it was increasingly
adopted. Indeed, this was one of the outstanding trends in pre-
fabrication during the latter part of the thirties— the abandonment
of metals and of grandiose schemes which had come to nothing,
for the use of wood on a more modest scale in a way that involved
but a limited amount of prefabrication. Whether the metals would,
in the end, prove to be the most useful materials for the industrial
ized production of housing was for history to tell. Many held out
hopes for their success in the long run, but they had received at least
a temporary setback.

4. The Achievement

By 1940 there were not more than 30 firms 75 in existence which
were manufacturing and selling prefabricated houses on a steady
basis.78 The great bulk of the production was of a sort that in
volved comparatively little in the way of new materials or prefabri
cation: precut and panelized wood frame (dry- wall) construction
(see Figure 10). All in all, excluding the precut houses, not more
than 10,000 prefabricated units 77 were produced between 1935 and
1940, or less than 1% 78 of all the single-family homes built in non-
farm areas during that period.

75 The files of the Bemis Foundation would indicate this.

76 Among the more prominent of these were:

Adirondack Log Cabin Co., Inc., New York, N. Y.

Allied Housing Associates, Inc., Langhorne, Pa.

American Houses, Inc., New York, N. Y.

Crawford Corporation, Baton Rouge, La.

Ivon R. Ford, Inc., McDonough, N. Y.

General Houses, Inc., Chicago, 111.

Gunnison Housing Corporation, New Albany, Ind.

Harnischfeger Corporation, Port Washington, Wis.

E. F. Hodgson Co., Boston, Mass.

Houston Ready-Cut House Co., Houston, Tex.

National Homes Corporation, Lafayette, Ind.

Pease Woodwork Company, Inc., Cincinnati, O.

Southern Mill & Manufacturing Co., Tulsa, Okla.

Willisway Construction Co., Chicago, 111.

77 No precise figures could be found. This is the estimate of Miles L. Colean,
American Housing; Problems and Prospects (New York: Twentieth Century
Fund, 1944), p. 147.

™ Housing Statistics Handbook, p. 6.

5. Summary

These, then, were the characteristics of prefabrication in the
thirties: a huge amount of interest, but few houses; active participa
tion in various ways by non-commercial institutions, government
agencies, and the large corporations; a profusion of structural ideas
only a few of which were technically and economically sound; and
the failure of these to achieve real commercial success on a large
scale because no one had yet brought together enough intelligence
and capital to develop an integrated building organization whose
operations extended from the procurement of materials through
manufacturing to selling, financing, erecting, and servicing the
home.79 Among the firms which sold houses on a continuing basis
there were several noticeable traits. There had been a retreat from
steel to wood, and from flat roofs and battens to Cape Cod cottages.
On the average, more and more was being included in the house
package, though as yet few companies had gone beyond the shipping
of wall panels and either panelized or precut floor and roof mem
bers to the packaging of a complete house with all materials and
mechanical equipment. There was, furthermore, a very minimum of
prefmishing. And, in the field of distribution, there were at least
two emerging patterns, besides those of the firms which catered to
such specialized shelter needs as vacation cottages and oil field
dwellings. One was the dealer organization, exemplified by Gunni-
son, through which many dealers sold houses one at a time to a
customer at a time; the other was the array of contractors and opera
tive builders through which American Houses was selling its product
in large groups to an anonymous market.

G. The Analogy with the Automobile

One other characteristic of the thirties which deserves mention is
the tendency to draw an analogy with the automobile industry; it is
one that has dogged prefabrication throughout most of its history,
but had an especially large influence in shaping the theories of this
period. Writers never tired of pointing out the example set by
Detroit in the mass production and mass distribution of automobiles
and never ceased lamenting the fact that the housebuilders had not

79 For an outline of reasons for failures in prefabrication, see John E. Burchard,
"How Better Houses WiU be Built," Technology Review, 39 (July 1937), 415-20.

wrought similar miracles. But in their haste to draw the analogy and
in their impatience with the building industry, they often forgot to
consider important factors. An obvious one is the bulk and weight
of the product, which directly affect the optimum degree of factory
assembly and the optimum factory size and location.

1. Subsidies and Land

Perhaps not so obvious is the fact that the housing industry had
yet to receive any such subsidy as the transportation industry had
received— the canals and railroads in the form of land grants, and the
automobile industry in the form of roads and highways.80 Cer
tainly, transportation depended no more definitely upon cheap land
than did housing. A public program to plan land use and write
down high land costs for housing might therefore be defended as
was an equivalent subsidy. Indeed, such an opinion was expressed,
not by long-haired radicals, but by the business magazine, Fortune,
which in 1932 wrote: 81

The $2,000 house on the $2,000 lot is no answer to the demand for a
$4,000 dwelling. Obviously, however, industry cannot alone insure the
stability and initial economy of the land. The fact is that aside from
temporary conditions due to the present emergency, land in or near urban
centers where the housing need is greatest will not be available at the
price necessary. The human inclination to speculate in land values will
see to that. Nor will neighborhoods protect themselves (and their mort
gagees) against blight of their own accord. Only some form of govern
mental intervention can secure the ends desired. It is therefore obvious
to the merest selfish considerations of private profit that the housing manu
facturers must associate themselves in their land purchases and in the
planning of their houses on the land, with some organization having gov
ernmental powers to condemn land in the first place and governmental
powers to protect it afterwards.

Our concern is not the soundness of this argument, but merely the
issues it raises: first, unlike automobiles, houses are not complete
until they are placed on land, and the price of land is a part of the
total over which the manufacturer has no control; second, public

80 Before 1927 motor vehicle owners paid less in motor vehicle taxes than
"their share" of road and street costs (Automobile Facts and Figures, ed. 22
[Detroit: Automobile Manufacturers Association, 1940], pp. 48-9). The public
powers used in road construction are much more important in this respect, how
ever, than the financial assistance itself.

** Fortune, VI (July 1932), 107.

intervention and public subsidy have played a part in the indus
trialization of more than one aspect of the national economy, and in
comparing the industrialization of shelter with that of other com
modities, they should not be overlooked.

2. The Character of the Innovation

A third point often neglected is this: of the new products which
have successfully been mass produced and distributed, most have
either offered a quality or service that differed substantially from
that of existing objects intended for the same purpose, or else they
have succeeded in performing a service that no previous object pro
vided at all. The radio, the telephone, and the automobile are all
examples of the latter. True, there were means of communication
and transportation at the time of each of these inventions, but the
new service was radically different from any existing at the time.

Now consider the house. With one or two possible exceptions, no
prefabricated house has yet provided shelter services that a con
ventionally built house could not duplicate. The recent inventions
which have entered the home, such as sanitary facilities, electric
lighting, cooking and refrigeration equipment, convection and radiant
heating systems, air conditioning, dish- and clothes-washing ma
chines, all could be and have been a part of the house built at the site
by traditional methods. For a long time, conventional houses have
been able to provide perfectly adequate shelter; they have been
strong enough, have protected the human body from the elements,
and have enabled it to maintain a comfortable temperature in an
environment which is visually and acoustically satisfactory. The
chief problem has been the economic one of providing the mass of
the population with such housing. Special factors occasionally favor
the prefabricator— remoteness of the site, importance of saving time
or labor at the site, or speed in selling, for example; but these are un
usual. Thus he has had to offer a product whose principal reason
for being purchased was that it was claimed to be cheaper than a
house of the same quality built by traditional methods.

This is not at all the situation which faced the early automobile
manufacturers.82 No horse, no matter how high priced, could do

82 Mark Adams, "The Automobile; A Luxury Becomes a Necessity," in Walton
Hamilton et al, Price and Price Policies (New York: McGraw-Hill, 1938), pp.
27-81.

what a car could. Conversely, there were many people who would
buy a car, even at a cost of several thousand dollars, simply because
it would provide excitement and exclusiveness. For more than ten
years the automobile manufacturers exploited their product's role
as a luxury commodity and sought to give cars more weight, power,
comfort, and brass ornaments to meet the demands of those who
could afford them. It was not until Henry Ford began his sys
tematic move to reduce prices in 1907 that the trend was reversed,
and even then Ford had no competitors for a number of years. Dur
ing the period of greatest expansion in the automobile industry the
demand for the services that only a car could give— immediately
available mechanized transport, speed, convenience, a new freedom,
and a new mark of prestige— was so great that firms were able to
finance their expansion largely by requiring deposits from their
dealers in prepayment for deliveries. Later the industry was in a
position to use its profits for expansion and had no great need for
recourse to the banks and the stock market. What a different course
has marked the initial stages of the industrialization of housing!

3. The Question of Durability

At least one other factor ought to be considered in comparing the
mass production of houses with that of other goods: the life of the
product. From the great durability of housing arises the circum
stance that the prefabricator must compete not only with the con
ventional builder but also with the vast supply of existing housing,
which at any time far exceeds the annual production. From this
great durability also stems the feast and famine character of house
building, with which both those in the industry and those who
would enter to revolutionize it must reckon. We may remember
that the transition from a situation in which demand represented
primarily first purchases to that in which demand is chiefly for
purposes of replacement was achieved fairly smoothly in the auto
mobile and radio industries because of the relatively short life of the
product. Through a combination of technological obsolescence, style
obsolescence, and physical depreciation, the average life of the auto
mobile was established at about nine years 83 and that of the radio

83 C. F. Roos and Victor Von Szeliski, "Factors Governing Changes in Do
mestic Automobile Demand," The Dynamics of Automobile Demand (New York:
General Motors Corporation, 1939), p. 48.

at about seven years.84 Compare this with ;the life of the house,
which is often roughly estimated at 70 years, but which in many
cases is known to exceed 100 years. Suppose it were possible to
rehouse most of America with manufactured homes; would the
industrialized housing industry then be able to readjust itself to
producing primarily for a replacement demand? Could it reduce
the life of the house and still produce a salable product? What
about the cultural values that a house symbolizes— do these affect
the nature of a product which has a long history and tradition and
is not just the child of an advanced technology? What about the
land to which houses must be attached, and the ultraconservative
branches of finance and law which deal with land?

These and other questions have to be considered in attempting
an analogy between the mass production of houses and of automo
biles. Any comparison which neglects such problems can at best be
superficial.

V. 1940-1945: The War Period

Just as the prefabrication industry was struggling to get on its
feet, the defense housing program hit it and knocked it off balance
with a whole new set of problems. Instead of a future of slow
development through concentration on key areas of difficulty, such
as distribution, prefabricators were faced with the prospect of a
huge market or practically none, depending on whether or not the
federal agencies 85 in charge of the war housing program could be
convinced of the industry's capacity to do a major part of the job.

84 Julius Weinberger, "Basic Economic Trends in the Radio Industry," Pro
ceedings of the Institute of Radio Engineers, 27 (November 1939), 708.

85 The War and Navy Departments, the Maritime Commission, the Federal
Works Agency— which had inherited the United States Housing Authority and
the Public Buildings Administration— the Reconstruction Finance Corporation,
and the Farm Security Administration were all active in the new construction
aspects of the defense housing program during its early history. On February
24, 1942, the President ordered the establishment of the National Housing Agency
which took over the housing functions of 16 non-military government agencies
and units.

A. Prefabrication on Trial

In the early days of the program there was more than a little
skepticism expressed in government quarters. The failures of pre-
fabrication in the preceding decade remained more firmly in mind
than the occasional successes, and certainly instances of the latter
sort had been of a modest rather than an overwhelming nature. But
prefabrication ofFered potential advantages, and after a good deal of
investigation and debate one of the agencies, the Public Buildings
Administration, arranged a demonstration project at Indian Head,
Md., where the prefabricators were to show what they could do.
In the spring of 1941 contracts for 650 units went to 11 companies,86
several of them without previous experience, which were to compete
in the erection of demountable houses on sites provided by PBA.
The systems represented were chiefly of plywood, insulating board,
or plasterboard on wood frame, although in spite of the materials
situation two firms used steel for the exterior wall. As a demonstra
tion project, Indian Head was not a great success. By the time
it was actually under construction some 13,000 prefabricated houses
were already being built or on order.87 Furthermore, the perform
ance of the inexperienced firms was not a credit to the industry as a
whole, and the problems at the site, particularly in the joining, fitting,
and alignment of panels, served more to show the prefabricators
their own weaknesses than to disprove the case for prefabrication in
general. It was shown, however, that the prefabricators could pro
duce at a price competitive with that of the conventionally built
house, especially when salvageability was taken into account. An
average performance in demounting the house, transporting it 40
miles, and reassembling it cost $474 and showed that on a dollar
basis 95% of the house was recoverable.88 Some of the actual projects
which got under way before the demonstration houses at Indian
Head had been started were more successful. A large project in
Vallejo, Calif., of about 1,000 Homasote and 700 plywood houses
was one of the best and clearly demonstrated the possibilities of

88 Allied Housing Associates, Inc., General Fabricators, Inc., Harnischfeger
Corporation, E. F. Hauserman Company, Home Building Corp., Lockwall Houses,
Inc., National Homes Corporation, Sears, Roebuck and Co. with General Houses,
Inc., Standard Houses Corp., Tennessee Coal, Iron & Railroad Co.

*iThe Architectural Forum, 75 (August 1941), 107.

**The Architectural Forum, 75 (September 1941), 189 ff.

prefabrication in a big development.89 In 1941 prefabrication was
for the first time on a mass-production basis. More than 18,000
units were built, probably more than had been produced in the entire
preceding decade.

B. Factors Favorable to Prefabrication

Prefabrication was used in the war housing program principally
because of three requirements: speed, demountability, and the re
duction of on-site labor and congestion to a minimum. These re
quirements, arising out of special situations, did a good deal more
to bring prefabrication methods into the picture than did the pre
war performance of prefabricators in producing housing under nor
mal circumstances. Although the industry had produced thousands
of permanent homes in the thirties, it had not earned a reputation
for unusual ability in this field; there were no outstanding firms of
great achievement, and consequently the great bulk of permanent
housing built during the war period, whether privately or publicly
financed, was constructed at the site by conventional methods or by
site-fabrication techniques. When, however, there was need for
obtaining permanent housing very quickly near a war construction
job or in a locality short in materials or building labor and super
vision, prefabrication often was adopted. And as it became evident
that shifting needs would be encountered, a considerable volume
of demountable housing suitable for long-term use was built, most
of which was prefabricated. Prefabrication was also used in areas
where it was necessary to keep site labor to a minimum for security
reasons, such as at the atomic bomb projects. After mid-1942 almost
all publicly financed housing was of a temporary 90 type. Not only
was it thought that further migration into crowded production areas
would be, in most cases, of short duration, but it became evident that
as the shortage of materials and labor grew increasingly acute the
standards of the buildings would have to be lowered. This brought
about such developments as a decided lightening of the structure,

89 William W. Wurster, project engineer on this project, reserved the right to
use entirely experimental design and construction on 25 of these units. It should
be noted that the construction cost of these 25, built by three local contractors,
was under $2,850 each, as compared with more than $2,900 for the regular
project units.

90 To be distinguished from demountable housing.

the exterior use of sheet materials not suited to the weather, the
elimination of as much metal as possible, and the sacrifice of space.
It also brought about the extensive use of stressed skin (prefabri
cated) construction in an effort to save framing lumber and gave
impetus to dry-wall construction (much of which was also prefabri
cated), which was faster and spared critical materials.

C. Signs of Pre fabrication's Growth

There were a number of signs of the growth of prefabrication
during this period. One was the number of different types of build
ings to which the techniques were applied: warehouses, hangars,
two-story row housing, schools. Another was the amount and scope
of speculative thinking and controversy. On the technical front
the idea of the panelized versus the sectional house was being
discussed, along with some variations such as the folding house.
Notoriety attended a number of proposed designs: Martin Wagner's
igloo-shaped house of steel, Buckminster Fuller's cylindrical house
made out of a grain bin (see Figure 11), William Stout's folding
house, Wallace NefFs unprefabricated but interesting hemispherical
house made of concrete sprayed onto an inflated balloon (Figure
11), and the Palace Corporation's suitcase house (see Figure 12).
Even early in the war the postwar house was a favorite topic for
discussion, and a glance through the architectural and homeowners'
magazines of these years would show how extensively the ideas of
prefabricated closets, bathrooms, and mechanical cores had taken
hold.

The concept of overall modular design also had increasing ac
ceptance, in theory at least, and was reflected in such diverse plans
as those of the Federal Public Housing Authority, Homasote Co.,
Ratio Structures, and General Panel Corporation of New York. And
in the realm of distribution there was much speculation over the fu
ture pattern of the industry. When The Architectural Forum hypothe
sized that in an integrated building industry the prefabricator would
sell to large developer-builders,91 Foster Gunnison was quick to
reply that no such pattern could succeed and that the necessary
diversification of sales risk could be had only if the prefabricator

91 The Architectural Forum, 77 (October 1942), 79-80.

sold to a great many small dealers.92 Vaux Wilson announced a
plan to sell his Precision-Built Homes through the department
stores because of their vast experience in merchandising. Still an
other sign was the amount of interest in prefabrication shown by
organized labor. The CIO, which in the years just before the war
had made small inroads into the AFL's building industry territory,
talked in big terms of the industrialized production of housing and
the industrial form of unionism that would come with it. A good
many wartime prefabricators had CIO shops, but in spite of a lot
of conjecture about AFL-CIO conflicts in war housing there was
little actual trouble. In light of the postwar developments to date,
the CIO's invasion of the housebuilding industry seems to have been
largely a temporary affair.

Another sign of the industry's growth was the formation in 1942
of the Prefabricated Home Manufacturers' Association, set up to
disseminate information, establish industry standards, study distri
bution problems, improve manufacturing methods, make cost and
accounting studies, and serve as a medium for the exchange of ideas.93
Prefabricated Homes,9* a monthly trade journal similarly aimed at
giving the public a clear and favorable picture of the industry, first
appeared in April 1943, instigated at least in part by PHMA. In
September 1943 PHMA changed its name to the Prefabricated Home
Manufacturers' Institute and expanded to include 12 charter mem
bers, with Walter Ahrens of Southern Mill & Manufacturing Co. as
president. Among the first things for which PHMI fought were
certain changes in the FPHA plans on which, since September 1943,
all prefabricators had had to bid for war housing.

Other indications of growth in the industry were to be found in
the number of active firms and their figures on output. Neither of
these statistics is ever very precise because the lines between what
is and what is not prefabrication and who is and who is not an active
prefabricator are so hard to draw.95 Even so, a few such figures
will give a rough idea of the picture. As against not more than 30

92 In an open letter to The Architectural Forum for the Prefabricated Home
Manufacturers' Association, November 13, 1942, in the files of the Bemis
Foundation.

93 The National Association of Housing Manufacturers, representing a few
of the more unconventional and newer companies, was organized in February
1947. Both are discussed in greater detail in Part II.

94 From January 1948 until October 1949 the magazine was known as Pre
fabrication. It is no longer published.

95 Lists of active prefabricators prepared by different sources at about the
same time have differed by as much as 100%.

active firms in early 1940, there were at least 100 firms in production
by the end of 1941. 96 At that time a government investigating com
mittee which inspected 35 of these plants reported that there was a
"reasonable certainty" of obtaining 27,450 units from the factories
visited, in quantities of 100-2,000 each, within 90 days from the
time orders were placed.97 By April 1943 The Architectural Forum
could assert that "there are now well over a score of prefabrication
plants that have each manufactured more than a thousand houses,
and many of which are now fabricating at the rate of several hundred
a month/'98

D. The Contribution of Prefabrication

A final summary of wartime prefabrication might credit the indus
try with 200,000 units.99 Of these, 116,390 were publicly financed
under the Lanham Act; 10° 16,000 were exported under lend-lease; 101
some tens of thousands were built by the Army and Navy at atomic
energy centers and American and overseas bases; and a compara
tively minor portion went into privately financed housing.

But big as this 200,000 figure may have been to an infant industry,
it still represented but a small part of the total of approximately

Q6The Architectural Forum, 76 (February 1942), 83. A list of manufac
turers and systems published by the prefabrication subcommittee of the Central
Housing Committee on Research, Design and Construction in February 1942 in
cluded about 200 prefabricates.

»7 Ibid., p. 82.

98 The Architectural Forum, 78 (April 1943), 72.

»»The Architectural Forum, 84 (April 1946), 137. Fortune, XXXIII (April
1946), 127, uses the same figure, but probably obtained it from The Architectural
Forum.

100 This represents about 25% of the war housing built with Lanham Act
funds. Of these, 104,862 were family dwellings and 11,528 were portable shelter
units designed for family use. Of the family dwellings, 1,428 were permanent,
66,901 were demountable, and 36,533 were temporary. About 11% of the 104,862
family units were fabricated in off-site factories and about 23% in on-site shops.
All the portable shelter units were factory fabricated. Source: Housing and Home
Finance Agency, in a letter to the Bemis Foundation, March 1948.

101 Early in 1945 the FPHA, acting for lend-lease, contracted for 30,000 units.
When lend-lease terminated, 16,000 of these had been started or completed.
The rest were not produced. Great Britain received 8,600 on lend-lease; France
bought the balance of 7,400 from FPHA. Source: Office of International In
quiries, Housing and Home Finance Agency, in an interview, June 4, 1948.

1,600,000 war housing units provided by new construction.102 It is
true that prefabricators made a major contribution in supplying hous
ing quickly in a number of key areas and in meeting the require
ments of special circumstances, but in the overall picture it remains
a fact that by far the largest part of the war housing was built at the
site by various techniques ranging from the conventional to the very
advanced. Large projects made it possible to embrace many aspects
of mass production at the site, such as standardization, specialization
of labor, and highly planned scheduling of processes and material
flow. Such projects also encouraged the use of power tools, jigs,
conveyers, cranes, and other paraphernalia of factory production.
Viewed in this respect, the war probably did more in rationalizing
and improving the efficiency of on-site construction than it did for
fabrication techniques in the factory, and it has been contended by
some that, relatively, prefabrication was thus pushed back.

The credit for this progress in methods does not all belong with
the conventional building industry. In a number of ways the devel
opment of site-fabrication techniques relied upon similar techniques
used in the factory of the prefabricator who should, therefore, be
given some credit. This development also resulted in part from
the efforts of the government which did a considerable amount of
research on site fabrication with its own technical personnel and
educated a good many builders in the use of better methods. Yet
the contributions of the "established" prefabricators (as of 1940)
in "know how" were perhaps less valuable than their general knowl
edge of the building operation. This may be a sign of their weakness
at the beginning of the war period, for other firms with little or no
previous experience in prefabrication found it possible to enter the
field and to build quite as readily, quite as successfully, and quite as
profitably, as the established prefabricators. It is probably also an
illustration of the fact that emergency production for a single con
sumer—a government at war— requires a pattern of operations very
different from that suited to the private sale of houses in normal
times. To be sure, in industries other than housing persons with
no previous experience in the field were successful operators, notably,
for example, in shipbuilding. But it is hard to think of an industry
in which this was so markedly the case as in prefabrication. After
a decade or more of gestation, the industry had not arrived at the
point where it could make a really unique and major contribution

102 Housing Statistics Handbook, p. 162, Table A; and Public Housing; the
Work of the Federal Public Housing Authority (March 1946), p. 8.

to an important war problem. This reflected not so much the in
competence of the industry as the extreme complexity of the prob
lem and the relatively small scale of the effort with which it had
been attacked.

E. The Effect of the War on Prefabrication

The war had a very positive effect on prefabrication. For the
first time production operations were put on a really large-volume
basis (though not always a steady one). A good deal was learned
about design and manufacturing techniques. Many firms attained
strong financial positions, and many new enterprises entered the
field. These and the signs of the growth of prefabrication discussed
above point to the positive effects of the war period on the industry.

But no evaluation of the effects of the war on prefabrication would
be complete if it did not include the harmful as well as the beneficial.
While the war gave impetus to the growth of prefabrication, it pushed
productive capacity beyond the industry's ability to distribute through
any of the channels it had thus far established. It aggravated the
unbalance between the prefabricated ability to produce and to dis
tribute. Furthermore, it made the marketing problem more difficult
because it gave the public a bad impression of the product. Whereas
the prewar prefabricated house may have been suspect as an interest
ing freak, the postwar product was often stereotyped in the public
mind as a dreary shack. A consumer opinion poll conducted by the
Curtis Publishing Co. in August 1944 103 showed that while 74.5$ of
those interviewed had heard of prefabricated houses, only 17.2% of
these would consider buying one to live in all year round. The rea
son given most frequently by the potential homeowner for not buying
a prefabricated house was lack of strength. Obviously, lightness was
being confused with weakness, and speedy erection with short life.
Another question indicated the public confusion over the industry's
diverse marketing methods, probably more a reflection of the various
speculative writings on the subject than of the actual practices them
selves. When those interviewed were asked where they would go to
buy a prefabricated house, 56.7% said they did not know; 13.7% said
the manufacturer; 10.8% said mail-order house or department store;
8.8% said dealer-builder; and 5.2% said lumber yard. Some two years

103 Urban Housing Survey, Curtis Publishing Co. (Philadelphia, 1945).

later, a Fortune poll104 gave much the same results: 70% had heard
of prefabricated houses, but only 16% were interested in living in them.
Thirty-three per cent said they would buy them only if they could
get nothing else, and when this group was asked what it disliked
about prefabricated houses, the replies were:

Unsatisfactory construction (included "not substantial
enough," "not strong enough," "not permanent," "not
warm enough") 67.4%

Lack individuality 13.4%

Too small 4.6%

All other 18.4%

Don't know 9.6%

(Some gave more than one answer.)

Thus, in meeting the need for demountable and temporary houses
of the lightest kind of construction, the industry was given an addi
tional handicap to overcome in the way of public prejudice.

104 Fortune, XXXIII (April 1946), 275.

Part A •

Chapter

1946-1949

GREAT EXPECTATIONS
AND
DISAPPOINTMENTS

The preceding chapter outlined the development of prefabrication
from early efforts through the war housing program. This chapter
is devoted to a description of the industry during the few years since
the end of the war. If in the first postwar months the homebuying
public and much of the business world were overly enchanted by the
promises of prefabrication, they probably have recently been as
grossly disenchanted, so that now, in a number of areas at least, the
opinion is that prefabrication has been tried and found wanting; that
the issue is settled: "prefabrication isn't practical." This chapter
might well begin with a protest against too great a disillusionment.

I. Background

A. The Shortage

The background against which prefabrication played its role in
the early postwar years included, among other things, a house-hungry
public, some significant shifts in political opinion, a major building
boom, and, not unrelated, a major inflation. The nation had been
hearing about the postwar dream house for four years. On top of
a cumulative shortage growing through the thirties and a shortage
caused by the cessation of normal building during the war, there
were returning veterans and high marriage and birth rates to be
reckoned with. It was estimated that as many as 3,000,000 houses
would have to be built in 1946 and 1947 just to keep the situation
from becoming worse.1 Furthermore, the great bulk of these homes
had to be provided for families in the middle- and lower-income
groups. Many looked to prefabrication to meet a major part of this
need. It is true that in the public mind there remained a picture of
the minimum standards to which prefabricators, through no fault of
their own, had had to build during the war. But many also believed
that World War II had done for prefabrication what World War I
had done for the automobile industry. Dream houses would roll off
production lines by the million and somehow end up in suburban

1 Wilson W. Wyatt, Housing Expediter, Veterans' Emergency Housing Pro
gram; Report to the President (February 7, 1946), p. 4.

neighborhoods behind rose bushes and white picket fences. A group
of startling housing ideas paraded before the eyes of the reading and
movie-going public: the Dymaxion house, the Tournalayer, the "solar
house." 2 Houses would be built of wood, as in the past, but large
numbers would also be built of concrete, steel, aluminum, plastic-
impregnated paper, and many completely new materials.

B. The Wyatt Program

In such an atmosphere, Wilson Wyatt was summoned to Washing
ton by the President in January 1946 to become the Housing Expe
diter. Five weeks later he submitted a program to the President
establishing a goal of 2,700,000 housing starts by the end of 1947
and calling for "the same daring, determination, and hard-hitting
teamwork" with which the nation had "tackled the emergency job
of building the world's most powerful war machine." 3 Private enter
prise was to assume the leading role in this task with the aid of ex
tensive federal measures aimed at expanding and directing produc
tion. The labor force in residential construction was to be tripled,
and local voluntary committees were to be established to help veterans
find homes, eliminate building bottlenecks, provide sites, reform
building codes, and speed the housing job in general.

Most of Wyatt's legislative proposals were enacted by Congress in
May 1946 as the Veterans' Emergency Housing Act.4 The program
which emerged from this legislation set out to increase production
by using surplus war plants, by making premium payments to stimu
late manufacturers of materials, by guaranteeing markets for new
types of materials and prefabricated houses, and by the financing of
new enterprises through Reconstruction Finance Corporation loans.
It sought to direct materials flow by curbing non-residential construc
tion and establishing a system of priorities, allocations, and restric
tions on house size; and to check the strong inflationary tendencies
(which had been aggravated by the liberalized FHA-financing pro-

2 These terms are explained at length later in the book.

3 Wyatt, op. cit., p. 1.

4 Public Law 388, 79th Congress, approved May 22, 1946. Wyatt requested
as an essential part of his program passage of S. 1592, the Wagner-Ellender-Taft
bill. This was a comprehensive long-range legislation providing for increased
FHA insurance, public housing, urban redevelopment, and other measures.
Congress did not pass this bill, chiefly because of the public housing provisions.

visions of the act itself) by controlling prices of materials and finished
houses.

In short, the Housing Expediter and his executive powers with re
spect to other agencies such as the Office of Price Administration and
the Reconstruction Finance Corporation represented an extension of
wartime government controls into a postwar period of acute housing
shortage. As a part of the overall goal, 250,000 prefabricated houses
were to be started in 1946, and 600,000 in 1947. Although this
program was under attack from certain quarters even in its earliest
phases, there was nonetheless a short period during which it ap
peared that the kind of cooperation and self-restraint necessary to
success would in fact be forthcoming. Committees were organized
on the local level, labor leaders pledged full support, and a number
of large industrial enterprises were reported to be ready to go into
prefabrication: Henry J. Kaiser, Higgins Industries, Inc., Douglas
Aircraft Co., Inc., Beech Aircraft Corp., Consolidated Vultee Aircraft
Corporation. This lent the program a certain amount of prestige
and, together with a favorable press, tended to bolster it against
growing criticism from numerous elements in the building industry.
But with the return to peacetime activities and interests, public sup
port diminished; broad political attitudes changed; the press and or
ganized criticism cried out against "government intervention"; gen
eral price controls were weakened, then abandoned; and the housing
program was the next to go. After the November election had placed
the Republicans in control of Congress and after Wyatt had run into
considerable opposition from a few key men in the Administration,5
he felt that the program was not going to receive the necessary sup
port, and in early December he resigned as Housing Expediter. Ten
days later the President announced the end of most of the controls;
premium payments, materials allocations, curbs on non-residential
building, and price ceilings were abandoned. The market guarantees
and loans to prefabricators, however, were continued until the end
of 1947, as specified in the law.

In retrospect, it hardly seems possible to classify the Wyatt pro
gram as other than a failure. Perhaps it was doomed from the start
as a grandiose and somewhat visionary experiment. Building starts
did accelerate in the late summer, but the overall total for the year

5 Much of the dispute was over Wyatt's inability to secure RFC loans for
selected prefabricators. The final breakdown came over his failure to have the
Dodge war production plant in Chicago assigned to Lustron Corporation, plus
a large RFC loan.

was 776,000 units started, a good bit below the target of 950,000.6
Furthermore, completions dragged because of shortages in materials
and labor. Prefabricators produced a total of 37,200 units in 1946
and 37,400 in 1947 7— not a bad performance in view of the extent of
shortages and unfamiliar restrictions, but far short of the program's
ambitious goals. Of these totals only a small fraction can be attributed
to the government measures. By the end of 1948 it was reported
that of the 32 companies which had secured guaranteed market con
tracts or loan agreements through the RFC only six were in active
production.8

On the other hand, it should be pointed out that the program was
never really given a chance. By the time the administrative machinery
was working, hostility was so great that few if any positive results
could have been expected. Whether the reconversion and expansion
would have been faster without any government program at all, and
would at the same time have provided for medium- and low-cost
homes (as critics of the program claimed) is a question that must
remain unanswered. Many of the critics spoke from long experience
and good common sense. Nevertheless, some of the boldest, most
risky, and in the long run perhaps most significant ventures would
never have gotten under way without a stimulus from the govern
ment along the lines proposed by Wyatt. The lessons learned from
one such really industrialized house manufacturer as Lustron, even
if it should never reach its production goals, may prove to be worth
all the money spent by the government 9 and the temporary doubts
cast on prefabrication as a whole. This, too, is a question that re
mains unanswered.

6 Both figures are for permanent dwellings, conventional and prefabricated,
and do not include conversions and trailers (for which the 1946 target was
250,000).

7 PHMI Washington News Letter, January 30, 1948, p. 3. 1946 total by the
Office of the Housing Expediter; 1947 total based on figures estimated by PHMI
and submitted to the Bemis Foundation. The 1948 total has been estimated
by PHMI at 30,000 and 1949 total at 35,000.

s Business Week, 1006 (December 11, 1948), 25.

9 It was estimated by the Office of the Housing Expediter in June 1948 that
the guaranteed market program, at that time almost completely closed and
settled, would not result in a loss to the government of over $3,000,000. What
losses the RFC will take on the loans it has made, many of which are out
standing, is hard to say.

C. The Birth and Death of Firms

Meanwhile the ranks of the prefabricators had swollen rapidly so
that, by the end of 1946, 280 companies had received priority ratings
from the National Housing Agency— as against less than 100 firms in
the industry some two years before. This rapid expansion served to
emphasize the ease of entry into the industry, but it was also a re
flection of the fact that many so-called prefabricators were nothing
more than distributors of building materials who did a minimum of
work on the materials they handled in order to secure higher prices
under existing regulations. There were others who called themselves
prefabricators in order to obtain priorities for certain materials and
who hoped to obtain guaranteed market contracts and capital loans
to start them on their way with little or no risk capital of their own.
Because the established firms feared that their reputations and that
of the industry as a whole would be adversely affected by the
failures of the newcomers and the poor quality of their products,
they sought to protect themselves by adopting quality standards and
by attacking the program which had brought about this great
influx.

Unfortunately, most of these fears were well founded. High ex
pectations attracted new enterprises, and the new enterprises had a
high death rate. Many never got into production at all; many others
failed; some retired to more conventional phases of the building pat
tern. By the end of 1947 the number of active prefabricators was
again less than 100, and in the wake of the failures there had grown
a profound skepticism regarding all that went by the name of pre-
fabrication— especially in banking circles. This purging of the pre
fabricators was somewhat reminiscent of early years in the automo
bile industry, and, if the outcome is as healthy, there may still be
cause for optimism.

Among the new enterprises were several of real interest, set up to
rate along lines which in one or more aspects represented a greater
reak with conventional building than was made by the vast majority
prefabricators, and often facing great difficulties as a result. In

is group might be included those who worked with aluminum and
lastic-paper sandwich materials,10 with standardized, universally

10 Southern California Homes, Inc., was the only company in this category
come close to production.

adaptable modular panels,11 with the sectional house idea,12 with the
"solar house" idea,13 and perhaps, because the design of the house
was approached with the same freshness that has marked the recent
interior design of trains, ships, and aircraft, with the hemispherical
Fuller house.14

D. The Building Boom

All the above activity should be viewed against the general back
ground of a building boom which proceeded with at least customary
violence. The number of permanent non-farm dwelling units started
with 209,000 in 1945 and went to 670,500 in 1946; to 849,000 in 1947;
to 931,300 in 1948.15 Residential construction costs went from 143.7
in 1945 to 159.2 in 1948 to 193 in 1947 and to 214.7 in 1948 (1939 =
100 ).16 The emphasis was primarily on single-family residences for
sale, and many families which might have preferred to rent were
forced to buy in order to secure any housing at all. A flourishing gray
market in building materials imparted to the whole endeavor a bad
odor. There were some significantly successful efforts at producing
good low-cost housing, chiefly by a few big operative builders who
made news because of the efficiency of their large-scale operations,
but by and large the housebuilding industry seemed to function much
as before— especially in regard to its characteristic of increasing costs
with increasing output. With construction activity using the avail-

11 The idea of selling such panels as building elements to contractors was pro
posed at first by such companies as The HomeOla Corporation and General Panel
of New York, but in both cases was later subordinated to the merchandising of a
complete house or houses.

12 In this country the Reliance house and the Prenco house ( produced by
Robert F. Johnson & Associates, formerly Prefabrication Engineering Co. ) and in
Great Britain the AIROH house remain in this category. Reliance is now in
production, and Johnson has abandoned stressed skin construction for standard
framing, sheathing, and siding in the conventional manner. 54,000 AIROH
houses were produced for the British Temporary Housing Program, and an addi
tional 15,000 have been ordered by the government as temporary housing.

13 Green's Ready-Built, which pushed this idea in 1946 and 1947, is now
defunct.

14 Fuller Houses, Inc., now defunct.

15 Includes privately and publicly financed units, prefabricated and conven
tional. Data from Construction, U. S. Bureau of Labor Statistics (April 1949),
p. 5.

16 Compiled, from figures of E. H. Boeckh and Associates, by NHA and HHFA,
Housing Statistics (April 1949), p. 8.

able resources to their limit, building costs were bid up by a flood
of purchasing power created by wartime savings, high incomes, and,
most of all, by easy mortgage credit. Between the end of 1945 and
mid-1948 the mortgage debt on one- to four-family residences had
risen about 65%, while non-farm family incomes had increased only
about 25% and the number of dwelling units by less than 10%.17 In
1947 more than half of the mortgage lending was being sponsored by
the federal government under legislation enacted by Congress (the
Veterans' Administration and Federal Housing Administration pro
grams ) and was on a basis that required a very minimum of builder's
or buyer's equity,18 or no equity at all. Factors such as these led the
Chairman of the Board of Governors of the Federal Reserve System
to describe "excessively easy mortgage credit" as "perhaps the most
inflationary single factor in the present [November 1947] situation." 19

By mid- 1948 a change in the situation was apparent. Mortgage
lending, especially on small homes, was tightening up. The trend
had been indicated for some months, and the expiration in April of
the liberal Title VI of the FHA program brought the situation to a
head. The re-enactment of a revised Title VI in the Housing Act of
1948, 20 passed at a special session of Congress in August, again made
very liberal government-insured credit available only to the lowest-
cost houses. A PHMI survey of its membership in the fall found
that nearly seven out of 10 placed the financing problem uppermost
of all the factors limiting their sales.21 Prospective homebuyers were
having trouble making the higher down payments, and banks were
slowing down their lending programs. Other housebuilders reported
the same difficulty. The lid was being clamped down. In the short
run, building activity might fall off until costs were shaken down, but
the very high proportion of the outstanding mortgage debt which had
been based on high prices would almost certainly be a serious con
sideration in the longer-range aspects of economic stability.

It might seem that, in a situation marked by an acute housing
shortage and an abundance of purchasing power, sales would present
no problem and the prefabricators would have succeeded in selling
more than approximately 37,000 houses per year in 1946 and 1947 or
30,000 houses in 1948. These figures represent 6.3% of all the single-

17 "The Economic Situation at Midyear," The Economic Reports of the Presi
dent, 1949 ed. (New York: Harcourt, Brace, 1949), p. 277.

18Marriner S. Eccles, "Inflationary Aspects of Housing Finance," statement
before the Joint Committee on the Economic Report, Special Session of Congress,
November 25, 1947, Federal Reserve Bulletin, No. 33 (December 1947), 1463-5.

19 Ibid., p. 1463.

20 Public Law 901, 80th Congress, approved August 10, 1948.

21 PHMI Washington News Letter, October 8, 1948, p. 1.

family dwelling units started in 1946, 5% of the total in 1947, and 3.9%
of the total in 1948. 22 But the fact that the market was there was not
enough— it had to be reached; the marketing process had to be or
ganized, and for a number of reasons which are outlined later in this
chapter this was perhaps the central problem of prefabricators in the
period of our study.

II. The Prefabricator: A Stage in Industrialization

The condition of the industry in the early postwar period is sum
marized in this section by grouping the many different prefabrication
operations according to the degree of their industrialization, under
headings which represent the major categories into which the indus
try may be readily divided.

A. The Panelized Wood Frame House

Least industrialized and most typical are those prefabrication enter
prises which have brought the production of a standard wood frame
structure into a shop where modular panels or room-size panels are
fabricated from lumber studs, sheathing of lumber, plywood, or some
type of wallboard, and insulation. Exterior and interior wall sur
face materials are applied in either shop or field with about equal
frequency. Such a shop is equipped with jigs, power saws, planers,
jointers, and other woodworking machinery, and perhaps some type
of machine to simplify nailing. There is a minimum of factory work
on ceilings, roofs, and floors, usually amounting to not more than
precutting. The house package which is shipped to the site repre
sents somewhat less than half of the cost of the finished house, less
lot. Since the design of the structure is quite conventional (see Fig
ure 3), except for its panelization, and since the tools used are very
largely the same ones that might be found in the precutting section of

22 590,000 single-family dwelling units were started in 1946, 740,200 in 1947
(Housing Statistics, Housing and Home Finance Agency [March 1949], p. 2),
and 766,600 in 1948 (Bureau of Labor Statistics).

2"x4" plate

Gypsum board
Interior finish usually
applied in field

2"x4u studs
Typical spacing 16"
on centers

Insulation metal foil

Wood sheathing

Building paper
or equivalent

Wood siding
Exterior finish varies
with builder

2°x4' sill plate

Figure 3. A Typical Wood Frame Panel

a medium or large site-builder's operation, there are few opportunities
for cost reduction through saving of materials and greater labor pro
ductivity. The principal cost reductions come through working con
ditions which usually are more convenient, through better division
of labor and organization of the work, and through large-scale pur
chasing. Other potential advantages include better control of ma
terials, more standardized production, better design, and less time
and money devoted to site work. Such advantages over on-site con
struction quite naturally increase as the prefabricated volume in
creases or as the size of the site-built project decreases. Against
these advantages must be charged the costs of plant overhead, trans
portation, and marketing, so that when a final accounting is made
we find these prefabricators offering a house of about equal or perhaps
slightly better quality than the average site-built house at about the
same price, any difference in price depending on the number of site-
built and prefabricated houses in the projects subject to comparison
and the degree to which the marketing and manufacturing processes
have been correlated.

While the designs and production techniques of such prefabricators
do not represent any significant increases in overall efficiency, they
encounter a minimum of resistance in the distribution process. The
finished house is generally indistinguishable from the typical site-built
house in the lower- and middle-cost brackets. It has one or one and
a half stories, a pitched roof, clapboard or shingle exterior, walls
of the customary thickness, and other conventional features. Conse
quently there is little consumer prejudice against it. Furthermore,
it is very likely to conform with local building codes. The substantial
amount of trade done with local materials dealers and plumbing
and electrical contractors helps to avoid another source of resistance.

This is, of course, a gross description and within the group to which
it refers there are some wide variations in particular aspects of design,
production, or marketing. Yet the characterization applies with rea
sonable accuracy to at least half of the companies now active.23

23 The figure of 75 active producers was given as an estimate by the Prefabri
cated Home Manufacturers' Institute for 1948 (PHMI News Release, June 4,
1949). It is very difficult to determine this figure accurately, both because
entry into and withdrawal from the industry are relatively easy, and because
the line demarcating prefabrication from other manufacturing and building oper
ations is tenuous at best. For 1949, PMHI estimates indicate 85 companies in
business, all but three of which used wood for their principal material (PHMI
Washington News Letter, December 23, 1949, p. 1).

B. The Stressed Skin Plywood House

A greater degree of industrialization is achieved by the group of
prefabricators who produce stressed skin plywood panel designs (see
Figure 4). They have made several significant breaks with conven
tional construction practices. Most important probably are the sav
ings in the use of materials made through efficient design and precise
engineering. There is also a tendency to use the structural skin sur
face of plywood as a finish material as well, and to use such composite
materials as paper-overlaid plywood and various types of wallboard.
Other characteristics are the trend towards prefabricating more of
the floor, ceiling, and roof elements, and towards providing a greater
degree of prefinish than do the first group. The introduction of cer
tain factory techniques has resulted in some important labor savings.
In the factory of such a prefabricator, for instance, we should expect
to find many types of woodworking machinery, jigs, and probably
conveyer lines. We are apt to find glue spreaders, some type of hot
press for gluing, sanding machines, paint sprayers, and drying ap
paratus.

Unfortunately, the savings in labor and materials achieved by these
companies are countered to some extent by the resistance which is
frequently met in the local communities. Buyers may be unhappy
about the plain flat finish of painted plywood, or about the thin walls,
no matter how strong these may be in fact. Many of the prefabri
cators themselves feel that steps must be taken to conceal all joints,
on the theory that buyers dislike joints in their houses. It is not un
likely that the building code will contain some provision that ex
cludes, for instance, the type of wall construction; and since these
prefabricators have tended to supply more and more in their house
packages, they may run into some form of resistance from local ma
terials suppliers or local labor when they try to obtain certain goods
and services needed to complete the house. But while such obstacles
are very likely to be encountered in a region in which these pre
fabricators are, in effect, not known, as in New England, there are
large areas where their products have been widely accepted by con
sumers, building inspectors, bankers, and local building-trades people,
notably in the Midwest. In the numerous medium-size cities in this
region the houses of these manufacturers are competitive in price
with the lowest-cost housing being built and are apt to be somewhat
superior in such qualities of construction as structural strength and

2x3' framing
members
16* on centers

Plywood
Interior finish
applied in shop

Insulation batts

Vapor barrier
Insulation backing

Male Joint

Female joii

3/s plywood
Exterior finish
applied in sho

Figure 4. A Typical Stressed Skin Panel

workmanship. They have not yet been markedly lower in price and
consequently cannot be said to offer a solution to the problem of pro
viding new housing for families in the low-income brackets, although
the industry is now concentrating on reducing costs in every way on
special low-income models.

There are perhaps 20 or 25 prefabricators who are producing stressed
skin plywood houses, and although this is but slightly more than one-
quarter the number of firms in the industry, as a group they have
been producing between one-third and one-half the total number of
prefabricated houses sold in the last few years. In this group are
many, maybe even a majority, of the strongest companies— those who
have the best plants and the most extensive marketing organizations,
and are potentially most capable of conducting and utilizing technical
research.24

C. The Machine-Made Metal House

Last are those few firms which represent the most industrialized
segment of the field. As a group, if indeed they may be called a
group, they are much less homogeneous than are the manufacturers
of the two types of houses described above. And as a group they
have produced only a small fraction of the total number of prefabri
cated houses built thus far. Although, as the heading indicates, their
common characteristic is that they work principally with metals, by
no means all the metal house producers are industrialized enough to
belong in this category, and several of the largest firms in it concen
trate on farm, industrial, and utility buildings with dwellings as only
a minor part of their business.25

Because it represented the most completely industrialized of the
house manufacturers, the Lustron Corporation may be taken as an
example of this part of the industry (see Figure 13). Lustron was
long the subject of bitter controversy, not only because of the sub
stantial role of the federal government in financing it and helping it
to obtain materials, but also because of its use of porcelain enameled

24 For example: Crawford, Gunnison, Harnischfeger, Houston Ready-Cut, Na
tional Homes, Pease.

25 For instance: Butler Manufacturing Company, The Steelcraft Manufacturing
Company, Stran-Steel (a division of Great Lakes Steel Corporation), and Fenestra
(a division of Detroit Steel Products Company).

steel for both interiors and exteriors, and because it was by far the
largest and most heavily capitalized prefabrication venture to date.
What made this enterprise unique, in the last analysis, was its scale:
the extent of its resources in trained personnel, in plant and equip
ment, and in financial power. If its projected output of 100 houses
a day, or 30,000-40,000 a year, could be attained, this would be
several times the volume of the largest peacetime builders. The
Lustron Corporation invested some $15,000,000 in the types of
tools and equipment that have long been employed in a number
of mass-production industries but that have remained foreign to hous
ing, such as large shears, presses, punches, welding machines, and
enameling ovens. The design, engineering, and sales organizations
were conceived on a similar scale. Inasmuch as size (in terms of
capital resources) has long been regarded by many observers as the
single characteristic most needed in a housebuilding enterprise if it
is to overcome the inefficiencies and obstacles besetting the many
aspects of the traditional industry, the discontinuity with previous
experience which Lustron represented in the matter of scale is of
considerable significance. Many have come to regard this venture
as a crucial test case for prefabrication, and its receivership will be
said to prove the folly of its basic concept. Yet its value as a test
case may be limited by the turn in the housing market since plans
were made, by the heavy commitment to one material and certain
production operations (which restrict freedom of design if changes
are to be made), and by the degree of attention which has been
focused on the RFC loans made to the company and their possible
economic, social, and political implications.

D. Other Types of Prefabrication

Besides these larger groups of prefabricators there are a few work
ing with composite sandwich materials such as Cemesto or aluminum-
surfaced paper-plastic honeycomb cores ( Southern California Homes )
(see Figure 21), with sectional house design (TVA and Reliance), or
with certain mechanized on-site processes, usually in connection with
concrete ( LeTourneau, Ibec, and Vacuum Concrete, Inc. ) . The manu
facturers of cabinets, storagewalls, doors, windows, stairs, chimneys,
and kitchen-bath utility cores also belong in the picture of prefabri
cation as a whole; and making mention of them here will serve to

emphasize the importance of learning to think not how many pre
fabricated houses are being built, but rather how much of the "aver
age" house is prefabricated.26

III. Broad Aspects of Prefabrication

A. Modular Coordination

Although prefabrication is here treated primarily as an industry
rather than as a general development, mention should be made of
some of the lines along which prefabrication as a broad movement
is growing. Modular coordination (see Figure 5) is such a line. At
first glance, there may seem only a distant relationship between pre
fabrication and the effort to coordinate the standard dimensions of
all building components so that they apply to any building that is laid
out on the 4" modular basis without cutting or altering at the site.
Yet it can be seen that if building materials and components were
manufactured in coordinated sizes and with provision for certain
standardized joints and constructions, they could be assembled with
relative ease and little waste into a wide variety of structures de
signed along modular principles. Even more important in its long-
range consequences, if all dimensions of all buildings were coordi
nated, many products now independently dimensioned, like kitchen
equipment, could be made to fit together, and many major assemblies
now rarely mass produced, like staircases, could be produced and
marketed in stock sizes in the manner of windows and doors. The
reduction of site work and the increase of factory work, involving at
least a partial shift of the building process from site to factory, are
the inevitable results of a successful program of modular coordina
tion and represent a trend in the direction of greater prefabrication.
The modular movement which started with the work of Albert Far-
well Bemis in the twenties and gained momentum in the thirties was
given added impetus in the war period not only from within the in-

26 A detailed discussion of the entire industry during this period is contained
in Part II.

4- 4" 4- 4" 4" 4' «• 4' 4' 4" 4' 4T V V V V 4' V V V * V V V V * 4' V V V V V

FILE

FRAMING

BLOCK

BRICK

WALLBOARD

Materials are produced to fit
multiples of basic module
and produce uniformity

Non-modular construction
requires cutting of bricks
which produces material
waste and high cost

Materials made in multiples
of 4" reduce cutting and
waste and can be fitted
together simply and orderly

Figure 5. The Principles of Modular Coordination

dustry 27 but from the government as well.28 Although it has not pro
ceeded as rapidly as it might have, because of the hesitancy of build
ing products manufacturers to incur the expense of changeover in
the presence of a seller's market, the movement is an inherently self-
accelerating one, and we may reasonably expect increasingly rapid
progress as time goes on. Today there are more than 600 firms pro
ducing modular structural clay products, masonry, wood windows,
steel windows, and glass block, and committees are currently working
on the modular design details of other products such as floors, kitchen
equipment, toilet partitions, and shower stalls.29

B. The Rationalization of On-Site Building

Another effort which has embraced certain aspects of prefabrica-
tion is the "industry-engineered house" program sponsored by the
National Retail Lumber Dealers Association and The Producers' Coun
cil and directed primarily at the builder of fewer than 10 houses per
year. The concept of modular coordination is basic to this program
in its designs and use of materials. Out of the time studies and cost
analysis of the sample houses built by the Small Homes Council of
the University of Illinois has come, also, the conclusion that definite
savings can be realized through the use of preassembled lightweight
roof trusses, making it possible to close in the house quickly with no
interior bearing partitions and with unbroken floor and ceiling finish.
The flooring, heating, plumbing, and electrical jobs can then be done

27 Project A62, sponsored jointly by American Standards Association, The
American Institute of Architects, and The Producers' Council, Inc., was begun
in 1939. It has been carried on with extensive technical assistance from the
Modular Service Association, a non-profit agency supported largely by the sons
of Albert Farwell Bemis.

28 During the war, modular coordination methods made a large contribution
to the success of the defense housing program, particularly in connection with
houses on the design of which Modular Service Association worked closely with
the Homasote Co. The Office of Technical Services of the Department of Com
merce contracted with the Modular Service Association in 1947 for research and
development along these lines. The Housing Act of 1948, Public Law 901, pro
vided the Housing and Home Finance Agency with $300,000 for development
and promotion of standardized building codes and standardized dimensions for
homebuilding materials and equipment.

29 "Modular Coordination," HHFA Technical Bulletin, no. 3 (March 1948),
p. 53.

more efficiently because the interior space is free from any obstruc
tions, and interior partitions can be framed and partly finished while
lying flat on the floor, and later tilted up into position.30 These are
perhaps simple methods, and the ideas are certainly not new; yet
they are instances of the type of influence that the movement towards
prefabrication is having on construction practices at the site. Most
big operative builders today not only do extensive precutting but
also a considerable degree of near-site shop fabrication of components
such as stairs, plumbing stacks, cabinets, storagewall units, and frame
assemblies for windows and doors. That some of these techniques
have been adopted by small builders as well is only further testimony
that the "conventional" builder of today is by no means using the
same methods that Noah did on the Ark, despite inferences to the
contrary which have had some currency.

IV. Prefabrication: Nature and Cost of the Product

To return now to the prefabricated house itself, it has been
widely said that, compared with the lowest-priced conventionally
built housing in the community, the prefabricator has been making a
slightly better product for about the same money. This is a generali
zation which is, of course, subject to exception. Certainly low-quality
prefabricated houses have been erected in the last few years by the
less responsible members of the industry, but on the whole careful
control of materials, factory precision of measurement and assembly,
and controlled factory working conditions have enabled the pre
fabricator to meet, if not surpass, the average small-home construc
tion standards. Not all prefabricators are producing for the lowest-
price brackets, however; at least one has built houses for as much as
$40,000 and others produce in the $15,000-$20,000 range. But by and
large most prefabricators have been and are today reaching for the
low-income market, which means for the prevailing two-bedroom
house a median selling price of roughly $8,000, completely erected and

30 Research Report on Construction Methods, Technical Series E2.1R, Small
Homes Council, University of Illinois, in cooperation with Office of Technical
Services, Department of Commerce, pp. 32-3.

finished but not including the cost of the lot.31 Naturally, in the
seller's market following the war, selling prices for both prefabricated
and conventional houses tended to relate more to what was offered
than to costs.

In analyzing costs, however, one must bear in mind that efficiencies
of quantity production can be realized in the field as well as in the
factory, and it can therefore be quite meaningless to make a com
parison between the cost of a single prefabricated house erected
on an isolated lot and the cost of a single site-built house in a project
of a thousand. So large a project affords opportunities for economies
in the procurement of materials and in the work of grading, installing
utilities, and laying the foundation, and the builder is able to achieve
to some degree the same type of division of labor and consequent
specialization that characterize line production in a factory. It is
much more meaningful to compare the costs of conventional and
prefabricated houses where both have been built singly, whether in
small or in large groups. For a one-house project the prefabricated
house will typically show some cost advantage, perhaps as much as
10-20%. As the size of the project increases, the cost advantage of
the prefabricator is apt to decrease and the nature of the so-called
"conventional" construction process will change, the site builder
adopting more and more of the techniques used by the prefabricator
until, in the very large projects of the operative builder, the pre
fabricator typically offers no cost advantages. The most efficient
housebuilding to date (as measured by cost per square foot) has
been done in such large projects. They have embraced varying
degrees of prefabrication, some builders doing the work in their
own shops near the site, other procuring a house package from a
prefabricator's plant as far as 300 miles away. The patterns of these
operative builders have almost always been worked out in terms of
wood, still our predominant housebuilding material.32

31 Such a figure is approximate, because of geographical variations and differ
ences in standards. For the typical "economy" house of two bedrooms and
768 sq. ft. of floor area, the median sales price among members of PHMI for the
completed house, less lot, was estimated as $7,000 in 1949 (PHMI Washington
News Letter, December 23, 1949, p. 1).

32 A notable exception recently was the Byrne Organization, Inc.'s Harundale
project near Baltimore in 1946-1947, where welded steel frames formed the
basis of a structure using other materials such as plaster, stucco, aluminum
clapboarding, and asphalt shingles. The expense of setting up near-site facilities
to prefabricate structural sections for these houses has been cited as a major
cause of the financial troubles which later plagued this project. See The Archi
tectural Forum, 90 (April 1949), 143 ff.

V. Prefabrication: Current Problems

That prefabrication has not yet brought about marked reductions
in the cost of housing and that it has thus far accounted for but
about 5% of postwar house construction have been causes for both
discouragement and disillusionment. It is said that, in spite of its
promise, prefabrication has not offered any solution to "the housing
problem," that it has utterly failed to realize its goals. Although
the goals which some have held were unrealistic, it may still be
asked why prefabrication has not been more successful in reducing
costs and (to the extent that this question is not included in the
preceding one) why it has not been more widely adopted. The
answers to questions like these should be approached only through
an understanding of the problems facing the prefabricator— prob
lems deriving not only from the technical and economic considera
tions inherent in any comparable industrial process, but also from
the complex character of the housing field within which the in
dustry operates.

A. Locus of Operations

Under present conditions, with the majority of prefabricators using
wood in a relatively conventional way, the practice in single-house
projects is to leave something like half the work (in terms of both
man-hours and value added) to be done at the site; in large projects
the site work is a much larger part of the total. Whether because
wood, used principally, is a material which can be processed with
relative ease in the field, or because prefabricated houses have often
recently been built in groups, or because engineering advances over
conventional construction have not usually been realized, more ex
tensive prefabrication seems simply not to be economically justified.

B. Marketing

Once the house has been designed and the production scheme
worked out, there are two vicious circles which frequently confront
the prefabricator:

Vicious Circle A. Though the design is superior to current prac
tice, from the point of view both of design and production, "people
like what they know" and do not like this design because it is new;
the banks consider the house too great a financial risk because of
the public reaction; without loans, few houses can be built; and the
design remains unknown and unaccepted.

Vicious Circle B. Low volume of production means high unit
cost; high unit cost means a small market; a small market means
low volume.

These situations are not novel; they occur in many other fields of
design and production, though seldom, if ever, in so acute a form.
But they serve to place the necessary emphasis on the fact that there
can be no mass production without mass marketing. This was
pointed out from time to time in the past, and today it is a truism.33
Yet in the frantic rush of postwar activity, and with materials short
ages a major preoccupation, only a few saw marketing as a problem
of any magnitude at all, let alone as their chief one. In the midst
of a severe housing shortage it was perhaps natural to underesti
mate the extent of the selling effort required and of the obstacles
which would be encountered. The history of the thirties should
have provided some lessons in this regard, but it was too easy to
ignore these in the light of the war experience and the other prob
lems of the immediate postwar situation. Now, at any rate, this
has changed, and the topics of advertising, sales, dealers, and in
terim and permanent financing are of major concern to most pre-
fabricators.

Most prefabricated houses are currently marketed through the
agency of dealer-erectors who combine the functions of selling the
house to the consumer, helping him to secure permanent financing,
erecting it at the site, and, often, servicing the finished home. There
are probably as many as 2,000 dealer-erectors, some being small
homebuilders who put up only a half-dozen houses a year, others

33 A case in point was the statement of the William H. Harman Corporation,
in its petition in bankruptcy, November 29, 1948: "We attribute the company's
failure to its inability to overcome the complexities of distribution and the
difficulties of financing sales and erection."

being large builder-developers who work in terms of large projects.
The choice and training of these dealers are of great importance to
the prefabricator, for they must be able to supply him with a steady
stream of orders on which to base his production, and they must
be able to carry out the erection and completion of the house at the
site with efficiency and dispatch; otherwise they will add in costs
whatever the prefabricator may have managed to save in the shop.
The prefabricator must train his dealers not only in the mechanics
of the erection process, but also in a whole series of other marketing
operations: the approach to homebuyers, building inspectors, lend
ing institutions, and occasionally irate neighbors; an idea of what
constitutes good site planning, and some notion of a "reasonable"
profit. While a low price is a potent factor in stimulating sales to
dealers, it is by no means the only one that must be present. Diffi
culties presented by codes, building officials, local materials dealers,
local labor, banks, and the FHA, plus consumer prejudice, are all
problems which must be overcome by patient effort on the part of
the prefabricator and his dealers. Bargain prices alone do not solve
them. Furthermore, there is nothing that requires the prefabricator's
cost savings to be passed on to the ultimate consumer.34 The high-
volume incentive of the manufacturer is not necessarily shared by
the dealer-erectors, many of whom operate speculatively and must
work hard to assemble land, develop it, and arrange for the many
construction operations on each house. In a favorable market,
charging what the traffic will bear may look to them like the best
policy, and in the inflationary situation following the war there has
often been a tendency for them to price the finished house at about
the same level as conventionally built houses in the area even if,
while still allowing a "reasonable" profit, they might have priced it
somewhat lower. Thus it is possible that prefabrication may do
better when the market enters a definitely deflationary phase, al
though other factors then complicate the situation. Just what is a

34 The same is generally true of the manufacturers of prefabricated building
components. For example, the Ingersoll Utility Unit, incorporating kitchen and
bathroom equipment and a central mechanical core, cost, when installed, about
the same as or slightly more than comparable equipment supplied through the
usual channels and assembled at the site. In spite of this, many of the units
were sold, partly because procurement was thus simplified, and partly because
the plumbing contractors who installed these units could do the job in less than
one-third of the time it took by conventional methods and could thus turn over
their capital more rapidly, taking a larger number of profits on their sales.
Nevertheless, the Ingersoll Utility Unit Division of Borg- Warner suspended opera
tions on June 30, 1949.

"reasonable profit" is of course hard to say, but inasmuch as many
dealer-erectors are construction people who have been partially
weaned away from conventional practice, they may hold ideas that
conflict with the mass-production concepts of the prefabricator. In
the long run the prefabricators will have to leave the dealers enough
margin for profit to attract the kind of ability that is needed for the
job.

A few firms have strong and well-disciplined dealer organizations
which erect their houses at a fixed price schedule. These com
panies have had the wisdom, endurance, and resources to develop
extensive dealer outlets and train them well. But sometimes the pre
fabricator has not developed a good marketing system, or he has
relied, as a temporary measure, on a few large projects to carry him
along, or both. At first glance this may seem more economical than
the investment in time and money which it takes to establish many
small dealers whose cumulative efforts supply the plant with a
steady stream of orders. It may also seem more economical be
cause certain efficiencies of scale can be achieved at the site. But
unless large projects form just a part of a prefabricated volume, or
unless he has, through an extended period, established relations with
a number of large project builders who operate steadily, he often
finds it difficult if not impossible to keep his plant running efficiently.
The big projects materialize slowly; when they do come through,
the prefabricator's procurement, production, and traffic departments
are placed under a strain to meet high, but temporary, production
requirements; and after this there is apt to be a slack period. Such
a pattern is more characteristic of a general contractor than of a
mass-production enterprise, but, unlike the contractor, the prefabri
cator has a considerable plant investment and a labor force to worry
about. For this reason there is increasing emphasis, especially among
those firms which practice a high degree of prefabrication, on de
veloping a distribution system geared to making many small sales
rather than a few big ones, thus diversifying sales risk and increas
ing the prospects for steady-volume operation.

C. Public Acceptance

A special aspect of the marketing problem has been the difficulty
sometimes encountered in securing public acceptance. Occasionally
there is a real prejudice against prefabrication which is not limited

to a generalized opposition to something new, but leads to action
not only by potential homebuyers themselves, but by the community
as well, through deed restrictions, pressure on building inspectors,
and the like. It stems chiefly from dislike of the minimum-standard
prefabricated dwellings built during the war emergency under gov
ernment contract. The bad reputation acquired in this way persists
in spite of the fact that the vast majority of prefabricated houses
built since the war compare favorably in every respect with con
ventional houses in the same price class. Moreover, the very large
number of prefabricated houses which have been financed under the
FHA have had to pass tests a good deal stiffer than those for most
conventional houses. As the public has become aware of this situa
tion its hostility has lessened, and today it is principally the houses
of unconventional materials such as steel and aluminum and those
of unconventional architectural appearance that are apt to arouse
suspicion and opposition, although many communities try to exclude
prefabricated houses simply because they are small and inexpensive
and therefore likely to give little aid in meeting local tax burdens.
In regard to appearance, there has been a strong tendency to make
the prefabricated house indistinguishable from the conventional house
and to abandon flat roofs and battens.

It is unfortunate that the general trend towards public acceptance
is retarded by occasional poor-quality products which act to rein
force latent prejudice. On the other hand, in reaching for the lowest-
cost market prefabricators have to make compromises with what the
public has come to consider, often wrongly, quality. Many a pre-
fabricator, building sound houses which make use of new construc
tion methods and as a result have light walls of thin cross section,
has found it desirable to avoid publicity during the erection process,
lest the house be considered flimsy.

D. Building Codes

Building codes have presented a serious obstacle to the growth of
prefabrication. One very inhibitory aspect of codes is their diversity.
They are so numerous and so non-uniform as to make it difficult if
not impossible to standardize certain items for mass production in a
factory. Plumbing codes are perhaps the outstanding offenders in
this respect and have discouraged a good many prefabricators from
attempting to manufacture plumbing assemblies. Another unneces-

sarily restrictive aspect of building codes is that they are generally
written in terms of specifications rather than of performance stand
ards, and that the specifications are in many respects outdated.
Some codes effectively exclude broad categories of construction,
such as dry wall, by indirection. Many exclude new and more effi
cient structural methods and materials which, from a performance
point of view, are perfectly adequate but which fail to meet the code
specifications: for instance, code provisions ordinarily require fram
ing dimensions in excess of those necessary or economical in stressed
skin plywood construction. There are some even more restrictive
code clauses such as those requiring certain types of field inspection,
and those providing for preference for local contractors and locally
manufactured materials.

An added problem arises from the fact that even an up-to-date,
state-wide, minimum-performance code will be of little use to the
mass producer if the local communities retain the power freely to
impose their own restrictions in excess of those called for by the
state code. The prefabricator, who wishes a large market area for a
standard product, needs state- wide maximums as well as minimums;
he needs protection against local code provisions which exceed those
required in the public interest. The code problem has received a
lot of publicity and a great deal of serious attention in the past few
years. Many cities and towns are rewriting their codes; others are
adding provisions permitting the testing and subsequent uniform
acceptance of new materials and structures; performance standards
are to some degree replacing specifications. Groups engaged in work
to standardize codes include several building officials conferences,
the National Bureau of Standards, and the Housing and Home
Finance Agency. Congress recently appropriated special funds for
this purpose.35 But such work moves slowly, and it will require a
great amount of time and effort to persuade thousands of local com
munities to adopt the same overall type of building regulation. In
the meantime prefabricators are managing, through the use of trial
or experimental houses and the accumulation of legal precedents, to
convince towns of the soundness of their structures, and they are
making progress in their own right. Some of the companies in the
Midwest producing stressed skin designs feel that in their area of
distribution the code problem is no longer a matter of major con
sequence.

35 The Housing Act of 1948 included funds for research in two fields, modular
coordination and building codes.

E. Local Trade and Labor

The prefabricate! occasionally encounters opposition from local
materials suppliers, contractors, and labor, who see their earnings
threatened by the prefabricator whose package represents materials
and labor imported from another community. This has led some
materials dealers to favor old customers during periods of shortage
(which is perhaps only a natural reaction), to decline credit, to
insist on tie-in sales, and to press for building-code provisions pro
tecting their interests. Plumbing, heating, and electrical contractors
have at times declined to make installations on equipment not fur
nished by them because in so doing they lost their customary markup.
Likewise, labor has from time to time refused to handle prefabricated
material, even when made by another local in the same brotherhood.
These obstacles have in general been of only minor and sporadic
consequence rather than a consistent source of trouble, but they add
to the prefabricated difficulties, and their net effect has been to
cause him to eliminate from his package items and work which he
might otherwise have included, often at considerable savings.

F. Financing

The aspect of marketing which has given most concern of late is
that of financing. After production is under way, the houses must be
sold and paid for, whether the sales are made to distributors, to
dealers, or direct to customers. Excepting in the case of the last,
arrangements must be made for some kind of interim financing. A
house package is an expensive item, amounting to between $3,000
and $4,000 in most cases, and few prefabricators are well enough
capitalized to extend credit until permanent financing has been ar
ranged on the house by a lending institution. Rather than tie up
his capital in this way, the house manufacturer must keep it turning
over in order to operate at high volume, and therefore he usually
asks the dealer to pay upon delivery of the package. This in turn
tends to put a strain on the dealer or to limit his volume, for he is
often unable to obtain the credit extended to builders by building
materials suppliers and must wait to receive payment from the bank
in installments as the house progresses, the first installment not being
paid, as a rule, until the house has been shelled in. A further com-

plication is the fact that a prefabricated house is a chattel and does
not become real estate until it is erected and attached to the land.
It is therefore subject to different laws and requires a somewhat dif
ferent credit instrument.

The acceptance corporation has been used in other fields to meet
this problem, and at least one company has succeeded in setting up
one to finance the sales of its prefabricated houses.36 When a house
package is delivered to a dealer, the acceptance corporation pays the
prefabricator for it and makes the first of several construction ad
vances to the dealer. The acceptance corporation subsequently sells
the mortgage, which represents final or consumer financing, to a
savings bank or insurance company. There has been considerable
interest in establishing independent companies to handle the prob
lems of a number of prefabricators in this way, combining chattel
and real estate mortgage financing, and possibly extending the scheme
to cover such items as refrigerators and ranges. To date, however,
no such independent company has appeared. The effect of federal
insurance of mortgages (FHA and VA) in decreasing the risk and
increasing the negotiability of mortgages as earning assets would be
very significant in any such development.

One other source of aid in facing interim financing problems has
been the extension of FHA insurance operations to cover working
capital loans to prefabricators and short-term financing of dealers.37

The last financial step is that of the ultimate buyer, usually in
seeking a mortgage from a bank. Though the banks have presented
no general obstacle, they have in some areas been very conservative
and very skeptical about prefabrication. Sometimes this conservatism
has made itself felt in the difficulty of obtaining working capital
loans; more often it has been exerted in the field of mortgage financ
ing. It is in no small sense true that the prefabricator sells his
house to the bank (or other mortgagee) rather than to the home
buyer. With present-day mortgages amortized over long periods,
usually considerably in excess of the average span of homeownership,
it is natural that lending institutions are concerned about resale value.
Their opposition to unconventional appearance affects site-built as

36 National Homes Acceptance Corporation, set up in 1947 by National Homes
Corporation with the backing of the American Bank and Trust Company of
Chicago, and later operating on RFC loans.

37 Under Section 609 of the National Housing Act, as amended. Under this
section commitments have been made for the insurance of loans to prefabricators
with provision for substitution of purchase contracts by the manufacturer, making
the principal amount in effect a revolving fund.

well as prefabricated houses, but, reflecting local prejudice, they have
sometimes objected to prefabrication as such, refusing to lend on it
or taking a mortgage for only a small fraction of the value. The
FHA has had a very important influence in encouraging banks to lend
on prefabricated houses. FHA standards have long been recognized
by the lending institutions, and when the Washington office of the
FHA approves a prefabrication system and issues an Engineering
Bulletin to that effect, there is much more confidence in these houses
on the part of the lenders. Inasmuch as 40-50% of the prefabricated
houses built in the last few years have been financed under the
FHA,38 it can be seen how important the FHA has been in the
general marketing picture.

Because of its almost decisive importance, a number of prefabri-
cators have seen fit to criticize the FHA, chiefly in circumstances
where they have had to make concessions in design or have had
approval flatly refused. It has been held that FHA standards are
too high, that they preclude the possibility of manufacturing a really
low-cost house, and that house standards should be lowered to the
point where homebuyers could afford them. This criticism seems
much less valid than those which have been leveled at various local
FHA offices for their conservatism in matters Of architectural design
and for structural requirements held to be unnecessary. Local offices
have at times declined to make commitments on houses approved by
the Washington office because of disapproval of such design features
as flat roofs and because of entirely local regulations regarding such
details as door widths. In some instances, different local regulations
were actually conflicting. This has led to suggestions that the
Washington office issue Bulletins approving the architecture of a
house as well as its engineering, thus forcing a certain amount of
conformance by local offices and giving the house a chance to be
accepted. Recently an assistant FHA commissioner has been ap
pointed with the function of assisting the prefabrication industry by
efforts to eliminate regional differences in rulings.

In the last analysis this and similar questions of policy rest on a
basic judgment as to just how much risk the FHA should take.
Many prefabricators feel that the best long-range likelihood of na
tional financial stability involves taking a certain amount of risk and
encouraging innovations to speed the development of better housing.

38 Estimate from PHMI survey of member companies, in a letter to the Bemis
Foundation, December 17, 1948.

VI. Conclusion

In the widest sense, current prefabrication is a growing movement
embracing a whole span of activities ranging from modular coordi
nation through the manufacture of various building components to
the production of houses themselves.39 Within this advancing front
there is distinguishable a house manufacturing industry which existed
in embryonic form through the thirties and is now a struggling,
growing infant. This infant industry produced more than 100,000
permanent homes in the three years following the end of the war,40
a small fraction, perhaps, of the total amount of housing built in
this period, but a significant total when one considers the investment
it represents. Although the prefabricator has not often been able to
produce at lower costs than the big operative builders working in
the great metropolitan areas, he has clearly demonstrated his ability
to compete with the lowest-cost housing produced in the smaller
urban areas where the operations of such large builders cannot be
continuously sustained. There has occurred a shaking-down process
in which the weakest firms have dropped out and the strongest firms
have grown stronger, their staffs expanded, their patterns of opera
tion crystallized. And there is a growing body of laws and institu
tions which are at least in part a manifestation of the strengthening
hand of the prefabricator: the extension of FHA operations, the con
tinuing aid of the government through the HHFA and the RFC,
the enjoining of certain malpractices through labor legislation, the
movement towards building-code reform, and the work of the Pre
fabricated Home Manufacturers' Institute and the National Associa
tion of Housing Manufacturers.

It has for some decades now been a paradox that the wealthiest,
most industrialized nation in the world should have been unable to
provide adequate housing for its citizens. Even though the house
building industry has moved to cheaper land and reduced the size of
its product, it has not been able to produce for the lower-income

39 "It has been estimated that about 20 per cent of the cost of the average
small conventionally built house can be accounted for in manufactured products
such as kitchen cabinets, kitchen and bathroom fixtures, heating plant, and the
like, as distinct from such materials as bricks, lumber and nails" (High Cost of
Housing, Report of the Joint Committee on Housing [Washington, 1948], p. 149).

*° PHMI Washington News Letter, October 1, 1948, p. 1.

groups. Prefabricators have earnestly sought to solve this problem,
almost always in terms of the free-standing single-family house, but
they have not yet come up with a solution. They have pushed the
lower edge of the housing market down a bit, but they have not yet
moved it a significant amount. What was in the thirties the problem
of the $2,500 house is now that of the $5,000 house; the problem re
mains essentially unchanged.

One suggested answer to the housing problem has been public
housing. Fearing public assumption of what has been a private
function, some prefabricators have called themselves "a bulwark
against socialized housing." But regardless of political views, it is
clear that, to date, prefabricators can at best claim for themselves
such a role only in terms of future potential.

6 Hodgson houses

7 A precut house of 1920, courtesy The Aladdin Company

_lL. .

•j&i

The Herford

ERE is a cleanly designed, substan
tial and altogether good-looking
dwelling. There is not one foot of

surplus lumber or timber, and yet the result
presented is pleasing and wholesome.

The Herford will accommodate a very large
family, there being four bed rooms and bath
on the second floor. And the living room and
dining room are proportioned to the needs of a
large family.

Every feature of this design will be found to
come under our plea for "modern, sanitary and
attractive" workmen's homes. And every feature is
planned at the same time to hold down the cost
The Charleroi Gas Coal Co., the Lincoln Gas Coal Co., and the Roa-
noke Mills Co. are among the many corporations who have found
housing satisfaction by the use of the Herford

The house requires but a 20x24 foot foundation and will take but a
25 foot lot. No expensive embellishments are to be observed. A
broad porch with the simple belt running around the middle of the
sidewalls relieve what might be extreme plainness.

No lower unit cost per person can Ixr secured than is |x>ssible in
constructing this house.

Price list attached gives our cost on this house.

SPECIFICATIONS

THE HERFORD
Size 20x24 Feet

«S2r2io^«^Jt££TS

•Mock* for ouuidc walla; Hcfeht of c

.nd pU.i.-r tor l.nlnc w.lU. «ilm«. .

8 Buckminster Fuller's first Dymaxion house

9 EarZt/ General Houses house

under construction

completed (1933)

1 Gunnison

2 National

Three

prefabricated
10 houses
of the
1940's

3 Crawford
( under
construction )

1 1 Two circular houses

Fullers "grain bin" house

Neff Airform House
(under construction)

1 basic package

2 half unfolded

3 unfolding roof

4 unfolding walls

12 A folding unit designed for emergency shelter,
the Palace Corporation

5 completed unit

73 Lustron houses

two-bedroom model, with garage

three-bedroom model

Part A •

Chapter

THE FUTURE

OF PREFABRICATION

I. Introduction

This chapter seeks to raise questions and to stimulate thought
more than to attempt prophecy. Yet one general forecast should
be made at the start, because it underlies much of the discussion
which follows. We believe that it will become increasingly difficult
to draw a line between prefabricated and conventional construction.
At the present time one-fifth of the average house is made up of
manufactured products rather than building materials in the ordi
nary sense. In the future more significance will attach to the degree
of prefabrication than to the numbers of prefabricated houses.

This does not mean that the house will become an exclusively site-
assembled product; development of the packaged house and the
sectional house will doubtless continue. The new processes and
procedures which typify many prefabricators seem sure, however,
to spread throughout the housing field and the construction industry
as a whole, and the benefits of mass production and mass distribu
tion will become generally available. In the end, the prefabrication
of houses may well prove to have been only a localized advance, a
specialized movement, in this general process of housing industriali
zation.

II. Current Trends within the Industry

Although many of the broadest problems facing prefabricators
have hardly begun to be understood, it is possible to point the prob
able future direction of trends visible within the industry in its pres
ent form. In large part, this discussion is based upon the detailed
analysis of the present industry contained in Part II; for that reason,
the comments made here may be very general in nature; and also for
that reason, the overall order of that Part will be followed.

A. Management

It has by now become abundantly clear that every step of the
prefabricated operations, from procurement through marketing,
exercises an important influence upon every other step. The process
used for erection affects the design as much as that used for produc
tion; mass sales depend as much upon good financing as upon good
design. In the future, therefore, the prefabricators will build up
balanced staffs of experts, or will retain consulting services, in order
to deal with this whole broad range of problems.

They will also take steps to develop large procurement, produc
tion, and marketing units in pursuit of the benefits of size. In the
period since the war, more than 30,000 houses have been sold each
year, although the number of companies involved has been sharply
reduced; this is roughly 30 times as many houses as had been sold
per year before the war. In the future, no doubt, there will remain
small specialized firms for special types of product or of market, but
the lion's share of the manufacturing business will go to fewer and
stronger companies. Much of this business will go to companies not
producing houses as such at all, but rather producing large compo
nents, either of houses or of buildings generally, for assembly either
at the site by individual architects, builders, and site developers, or
in fairly localized assembly plants.

At the same time, general industrialization of the building industry
will be in progress, a very noticeable element of which will be the
growth of dimensional coordination. This is a self -accelerating move
ment, and efforts now being made to educate manufacturers, sup
pliers, builders, and architects on the one hand, and the consuming
public on the other hand, seem certain to bear fruit.

As they grow in size, prefabricators will grow in responsibility and
in the desire to maintain a high sales volume over a period of years.
Long-range plans and policies and an understanding of the whole
housing market will become increasingly important, and a great deal
of attention will be paid to the devices of combinations, mergers, and
licensing agreements. In the past, there has frequently been specu
lation on the possibility of the development of a sort of "General
Motors" for the production of houses. We have described in earlier
chapters a few prior attempts to set this up; when the time seems
right, it will be attempted again.

Several of the prefabricators have already managed to integrate
their operations to some extent— more often by controlling or actually

100

owning their suppliers of basic raw materials than by controlling
their marketing operations. In the future, this process will continue
in both directions, seeking the advantages not only of simplification
in management and reduction in cost, but also of the marketing fea
ture of making it possible for the buyer to turn all his problems relat
ing to his house over to a single large organization. By the very
nature of the product and its marketing process, however, it seems
unlikely that there will ever be a housing "Detroit." The large pro
ducers will probably be enough different, one from another, to prefer
different production areas, and many advantages may be found to
lie in decentralized production.

B. Design

1. Materials

Wood. Since designs are very much a function of materials, the
future prospects of the use of different basic materials are of con
siderable interest. It has been said that the traditional domination
of wood as a housebuilding material is being threatened. In the form
of lumber, its use in prefabrication can be expected to decline.

Plywood and the related bonded paper ply materials, on the other
hand, seem certain for a while to maintain their popularity in the
prefabrication industry unless there is a substantial reduction in
supply and rise in price. Wood fiber products, already in wide use,
will continue to grow in popularity, however, and other products of
wood technology will continue their rapid development, some so dif
ferent from those of the present as to warrant calling them plastic
materials. This will almost certainly be the direction in which the
industrial use of wood for houses will turn.

Concrete. As for concrete, already an important building material
in the form of concrete block, many recent improvements in tech
nology will help to bring it into increased use for houses, but it can
not readily be made the object of mass distribution. Rather it will
come into its own in the site fabrication of huge projects, with in
creased reliance on mechanized, portable, and re-usable forming,
pouring, and curing equipment.

Clay Products. Structural clay products have been less widely

101

used in the building industry because of their increasing site-con
struction costs and the objections on thermal and acoustic grounds to
their use as a single-material wall. They will undoubtedly remain
an important element in the design of large site projects, but for fur
ther industrialization much will depend upon the success of current
research in improving their physical properties and in developing
larger and lighter units capable of production with greater precision.

Metals. The metals have a bright future, if there is to be an in
dustrialization of houses in the form of finish as well as of conven
tional framing. Steel is admirably suited to mass production, and the
major problems affecting its use in houses, condensation and corro
sion, are approaching satisfactory solution. The increasing produc
tion and decreasing costs of aluminum and magnesium make these
light metals very promising, and they have good properties for hous
ing purposes.

Plastics. Plastics, in the sense of materials molded under heat and
pressure, are already in use for trim and accessories about the house,
but because of their relatively high cost and low strength they have
thus far proved to be unsatisfactory basic building materials. Their
future lies in combination as binders and adhesives with other mate
rials such as wood wastes, wood fibers, wood veneers, paper, vege
table fibers, glass fibers, and the like, or as finish coatings.

It is the development of plastics along these lines which has made
possible the rapid development of plywood, and the plastic core ma
terials may in a few years become important building materials.

Wallboards. The trend towards dry-wall construction will con
tinue to spur the development of wallboards and composition boards
of various types, such as cement asbestos, fiber, and pulp, especially
when these are combined in sandwiches with other materials offering
different technical or finish qualities.

2. Large Panels

Related to the use of such new materials as the laminated wood-
plastic combinations and the metals is the trend towards large panels,
which avoid seam and joint problems by maintaining continuity of
surface, and simplify structure through a fusion of skeleton and skin.1

1For a good discussion of the principle of continuity, see Fitch, op. cit., pp.
183-5.

102

In this connection, there will be an increasing effort to prefabricate
those components of the house which offer large, unbroken surfaces,
such as the ceilings, roofs, floors, and partitions, whereas today the
major effort is directed at the walls. The inherent merit of frame
and curtain wall structures for many purposes and in many materials
will assure their continued development also, but the light, continu
ous, combined-purpose walls will advance more rapidly.

3. Factory Finishing

Along with larger continuous surfaces will come the development
of better factory finishing. In the metals this trend can be illustrated
by the vitreous enamel finish of Lustron. In the woods it may take
a new direction, such as resin impregnation or compression or both.
Albert G. H. Dietz points out 2 that the assembly of a frame and the
application of boarding offer much less opportunity for savings in
construction labor and time than do the finishing of floors, the paint
ing of woodwork, and the many other finishing details; and that
significant future advances may come from the use of impregnation,
compression, and high-frequency techniques to achieve the same
purposes.

4. Color and Texture

Of greater importance to most people than is generally recognized
are questions of color and texture, and here rapid developments seem
sure to take place. What the public considers high in quality is often
high only in finish quality, and manufacturing processes can produce
economical finishes of better performance and of greater variety in
color and texture than those now used in the housing field. At
present there is some experimentation with color, but the possibilities
of texture have gone almost unnoticed. It seems to be generally
assumed that people like uniform flat surfaces on their walls, and
to be further assumed by such companies as Lustron that they like
uniform color and washable finishes as well. Yet little is actually
known about the merits of different textures and finishes because in

2 "Progress in Wood Construction," Wood Preserving News, XXV (December
1947), HOff.

103

the past relatively few possibilities have been available for use in the
house.

It should prove to be desirable to produce surface finishes which
do not require constant cleaning, no matter how easily they can be
cleaned. Certainly this seems to have been the conclusion of the
makers of linoleum. A little texture— a fine corrugation or processed
pattern— together with an irregular color pattern might make it pos
sible to clean less often, and in addition add improved mechanical
and acoustical performance. Less uniformity should mean easier
production control, and corrugation or stamping should permit the
use of lighter gauges of metal. These possibilities are certain to be
explored in the future.

5. New Structural Forms

With the new materials and a higher degree of factory finishing
will come new structural systems and new plastic forms for the fin
ished house. Although in the mind of the typical homeowner the
house may be essentially rectilinear, there is plenty of historical prece
dent for other forms where the structural basis is other than post
and lintel. At present, consumer resistance to the Fuller or Neff
hemispherical houses would be violent but perhaps less widespread
than has been supposed. In one known instance, the majority of a
group of potential homeowners wanted to examine a hemispherical
house before making up their minds, although they had summarily
rejected a contemporary rectilinear design in favor of a traditional
design. An entirely new form may have a better chance for accept
ance than one close enough to a traditional stereotype to cause con
stant irritation because of its differences.

Revolutionary designers tend to feel that the logic of structural
efficiency has an overwhelming appeal. There seems little reason
to believe, however, that we demand a high degree of structural
efficiency in the house. Architectural design involves many prob
lems; and, in the future, basic considerations of plan will continue to
dictate the structure, rather than the reverse.

Indeed, the structure of the house is a mystery to the average per
son, and he rarely even shows an interest in it. This was illustrated
when a new type of steel construction was used in an exhibition
house put up in 1933 at the Century of Progress exposition in Chi
cago. Pleased with their achievement, the engineers responsible for

104

the design put glass insets in the walls to show construction features
and handed out questionnaires to find which aspects of the house had
most interested their visitors. Far above all else in terms of popular
interest was the presence of twin beds, and in second place by an
equally commanding margin was the use of Venetian blinds. The
construction was hardly mentioned. Clearly, the public expects
professionals and trained officials to watch out for its interests in
these matters.

6. Project "Variety"

The industry will gradually grow away from the tendency to seek
"variety" through the application of exterior materials, details, and
finish treatments to identical houses in the hope of giving the ap
pearance of that random collection of structures which has charac
terized our neighborhoods in the past. The results obtained by these
devices are rarely pleasant, and often they achieve only what William
W. Wurster has called "the monotony of slight variation." More
important in the future will be variation in color, in placement of
houses, in arrangement of the lot and street lines, and in relation
ships established with garages and other structures— a variation which
obtains its quality from a frank recognition of the basic similarity of
the houses involved.

It will be recognized that, beyond a certain size (the definition of
which requires study), a project of similar houses develops an op
pressive monotony which no artistry can dispel. Those living in
such projects know this, if the builders do not, because the reasons
lie as much in the formation of an oversized mass of similar family
groups as in the architectural effects.

7. Mechanical Cores

The mechanical services and equipment of the house represent from
about a third to as much as half of its production cost. It is certain
that the effort to design these as a unit core and to mass-produce such
units in ever larger components will continue. In the next few
years development here may come even more rapidly than in ra
tionalization of the rest of the structure. The difficulties now faced

by the makers of mechanical cores are certain to diminish, for im
provements along this line offer great cost savings, production and
erection simplifications, and sales and service advantages. One can
easily foresee the development of a mechanical core together with a
basic structural frame capable of carrying the weight of framing and
finishing the entire roof as well as a curtain of walls and windows;
and undoubtedly there will be special models of cores available to
provide different standards of service.

8. Integrated House

There will be developing at the same time the integrated house,
manufactured for sale as a single unit offering little or no design
variation, and incorporating all its mechanical apparatus. While this
may seem a logical extension of the mechanical core, it is in many
respects a quite different development, for it requires that the entire
house be dealt with as a unit, while mechanical cores may be used
in connection with "conventional" buildings or even existing build
ings, as well as with prefabricated houses.

C. Procurement

1. Materials

The obvious procurement problem is that of future supplies of
basic raw materials. For wood the situation at present is very dif
ferent from that of a few decades ago. Our forests have very rapidly
diminished, and although wood is the only one of the major raw
materials of building which can be replaced as a crop, not enough
concerted effort in that direction has yet been made in this coun
try. Yet for wood, and for plywood, it cannot be said that there is
any immediate prospect of a shortage.

Furthermore, there are new processes in operation and under de
velopment, making use of smaller pieces of wood in edge-grain ply
wood and employing special surface materials which permit the use
of smaller quantities and poorer grades of veneer. Illustrative of
the materials made by these processes are paper-overlaid plywood

106

(resin-impregnated paper bonded to rough plywood to give a smooth
hard surface), K-veneer (heavy kraft paper bonded to a single thick
veneer which has been slit and distended before bonding to increase
dimensional stability), and several types of wood core with bonded
metal surface.

There seem to be few procurement problems for the concretes and
clay products, and the story of the metals is widely known. The
supply of steel is a matter more of national policy than of the avail
ability of basic raw materials, although there may be significant
changes in the production centers and distribution systems in the
future with the development of new sources of ore and changing
price policies. As in the past, any defense emergency will mean the
pre-emption of steel supplies for war purposes, and the housing in
dustry will be forced to use substitutes to the fullest degree possible.

This is also true for the light metals, for which the future in terms
of raw materials and increasing production capacity looks very bright.
For both aluminum and magnesium, it is not the supply of raw
material so much as the cost of power which determines available
supply. Production of both metals increased greatly during the war
and is likely to increase again in the future. It has been estimated
that the aluminum used today in such elements as windows, insu
lation, roofing, and spandrels for the building industry is greater in
total amount than the prewar production for all purposes combined.3

2. Components

Where prefabrication amounts to little more than the assembly of
components fabricated by others, procurement obviously becomes
the heart of the operation, but many feel that one of the most im
portant contributions of even the typical prefabricator has been the
streamlining of building supplies and equipment distribution. This
function will expand in the future as supplies and equipment manu
facturers satisfy themselves of the reliability of the prefabricator as
a source of large and steady orders, and as further vertical integration
occurs within the prefabrication organizations themselves. Eventu
ally, more and more prefabricators will strive for the position common
in the automobile industry, in which the company is large enough to
control its suppliers.

3 Howard T. Fisher, "Prefabrication; What Does it Mean to the Architect?"
Journal of The American Institute of Architects, X (November 1948), 220.

1017

D. Production

1. New Processes

It is not possible to give general consideration here to the future
of industrial techniques in prefabrication industries. Processes which
are likely to show an increasing development in the next few years
may be listed, however. For wood, they include gluing instead of
nailing (a manifestation of the tendency towards more continuous
surfaces), high-frequency induction heating for the curing of glues,
and thermopressure molding of plywoods. For steel and for the light
metals it seems certain that there will be more common use of fac
tory finishes, such as vitreous enameling or some of the other forms
of baked finishes at present used for automobiles. The cellular and
corrugated core materials will leap into prominence with the devel
opment of any method of continuous strip production of the cores,
but this is not an easy problem.

2. Production versus Erection Economies

In the past, a great deal more energy in the design of the prefabri
cated house has been devoted to securing economies in the factory
than in the field, and often the result has been that unexpected field
costs have overbalanced the factory savings which were so carefully
planned. This will be discussed in more detail under Erection
(p. Ill), but it should be pointed out here that this lesson is being
learned, and that production schemes in the future will take into
account the efficiency of the operation as a whole.

3. Standardization versus Specialization

In any production scheme, attention must be devoted to the ques
tion of standardization. It is often said that parts should be stand
ardized and made interchangeable to the fullest degree possible, but
this depends very greatly upon the expected rate of production and
the variation in production models. If a single product is to be fabri-

108

cated in large quantity, there may be savings in designing specialized
parts for maximum efficiency without full standardization. To stand
ardize could be to make certain parts unnecessarily strong and thus
wasteful in materials. On the other hand, when parts are standard
ized as fully as possible, there may be greater simplicity in procure
ment, production, packaging, and erection. The prefabrication or
ganizations of the future will be better able to determine accurate
costs and to decide these production problems on a realistic basis.

4. Operational Decisions

Production operations themselves will receive considerable study:
the breakdown of the job into simple and repetitive operations, the
use of a continuous-flow production line to pace production, and the
use of jigs, of work-simplification and production-control techniques,
and of sound accounting procedures.

Many of the most important production decisions will depend
upon the expected market. Analysis of the market in some cases
will call for the decision to stay with wood construction and repeti
tive station operation, in order to permit considerable variation in
rate of production without undue plant costs. Once the choice of
steel is made, however, a production line seems indicated, and a
mass -marketing mechanism at a high level of stability is required.
No mechanism has yet reached this high level, but attempts will be
continued in the future.

In any case, it seems certain that the prefabricators will be among
the first to make available to the mass-housing market the new ma
terials and methods of construction, and many of the new items of
special service and appeal, in so far as these are well adapted to
factory production methods.4 As other builders fall in line with pub
lic demand for this development, the fabrication of an ever-growing
portion of the house will be transferred to the factory, until eventu
ally the operations of the entire housing industry will become so
advanced that little significance will remain in a distinction between
the prefabricators and the other producers of housing and building
components.

4 Factory construction itself has long been a proving ground for new design
ideas. See Fitch, op. cit., pp. 68-9.

109

E. Marketing

Throughout this book, emphasis has been placed upon the im
portance to prefabricators of building sound marketing organizations,
partly because mass marketing is always a prerequisite to mass pro
duction, but largely because uniquely difficult marketing problems
are presented by houses. Considered simply as physical products,
they involve great difficulties of assembly, packaging, transportation,
and, in most cases, erection, and the design and production processes
are intimately concerned with the schemes developed for overcoming
these difficulties. Briefly highlighted here will be a number of special
aspects of this general problem, to which prefabricators will give
increased attention in the future.

1. Packing

Prefabricated houses are generally transported by means of tractor-
trailer trucks, the trailer units sometimes being used also as movable
parts of the assembly line. In other instances components are pal
letized for easy handling, and in still others loading from component
bins is worked out as required by each order. In the future, if the
design emphasizes many standardized parts, there may follow a de
velopment of shipping containers for these parts which will them
selves become a part of the final house; and, a small but important
point, factory packing and loading will be designed for easy off
loading at the site, where usually there will not be available the
specialized equipment common at the factory.

2. Transportation

Generally it will not be possible to transport completed houses,
and that fact in itself offers a possibility for variety in the finished
house. Sectional houses made up in units of size suitable for ship
ment in trailers or railroad cars may be varied and combined in dif
ferent ways at the site to produce houses which are substantially
different one from another, and not merely slight external variations.
On the other hand, collapsing and folding houses are already well

110

known and may develop rapidly in the future. Especially when
they are made of the new materials with large continuous surfaces
and of lightweight construction, they offer a very good solution for
the problems of assembly, packaging, transporting, off-loading, and
erection. They also have the advantage of immediate roofing-in at
the site, which provides protection against the weather and permits
the prompt departure and re-use of the trailers. They do not ease
the problems of trailer size, of road loading and bulk, or of access to
the site, however, and in some cases these will offer serious difficulties.

3. Erection

In theory one of the great savings of prefabrication lies in simple,
expert erection handled by trained dealer organizations. In fact, of
course, such trained organizations have been the exception rather
than the rule because of the rapid growth of the industry and because
a new type of man is required in the dealer role. The old-time lumber
yards and the conventional builders in many cases appear to be un
able to reach full speed or efficiency in handling these new respon
sibilities. As a result, many a prefabricator is taking steps to create
his own dealers by training young college graduates in his plant and
later sending them out in the field and financing them until they get
on their own feet. The establishment of expert dealer organizations
will take time. As they come into existence, however, they will
bring about cost savings very rarely achieved up to now.

It should be added on the subject of erection that the small stand
ardized parts, or components, which have so many advantages else
where in a pattern of operations, tend to be at a disadvantage when
it comes to assembling them at the site, often in positions awkward
for manual labor, and to sealing the numerous joints that necessarily
are involved. The prefabricator of the future will be wary of the use
of such parts, particularly if they require extra strength or extra labor
to make them easier to handle at the site. On the other hand, the
smaller parts used in the erection process, such as bolts, screws, and
the like, will become as fully interchangeable as possible so that
time need not be wasted finding the right piece or trying to make
the wrong piece fit.

Ill

4. Regional Distributors

At present, few prefabricators make use of distributors in their
distribution channels, but the likelihood is that more will do so in
the future. When mass-production quantities reach into many thou
sands per year, it may well prove more efficient to divide the sales
area into several regions, preferably having common conditions of
climate and local design preference, and to ship the houses by effi
cient railroad or comparable mass transportation to distribution points
in these regions. It will not make sense to send 10 or 20 trucks per
day over the same basic route for hundreds of miles from the factory
before branching off to the local destination. Furthermore, regional
distributors offer a partial compromise on the issue of factory versus
site assembly. Site assembly usually means difficult and inefficient
conditions. Factory assembly, on the other hand, usually is a space
and overhead consumer, particularly when delay in the sales or ship
ment process requires stockpiling on factory floor space. For a high-
production factory, if knocked-down packages were shipped to re
gional distributors somewhat in advance of normal sales, these men
might perform a minimum of preassembly, and if orders were slow,
continue with the preassembly process as far as possible within the
limitations of the final means of transportation. At the same time,
they might be responsible for carrying out regional variations in the
basic house, along certain standardized lines. In cold climates they
might install more wall and fewer window units, extra insulation,
and larger heating systems. If regional construction requirements
varied significantly from nation-wide standards, certain standard sub
stitutions could be made at this point, for example in the plumbing.

5. Simplified Selling

One of the great advantages which the prefabricator can offer is
the simplification of the various steps through which the individual
purchaser must go in order to buy a house. This should start with
the establishment of a fixed price. In the future, prefabricators
will not continue to allow dealers to establish prices in their own
locale. The stronger firms already have their dealers quoting prices
from a fixed schedule under their control, and those firms will do
best in the leaner days ahead which can advertise the ' price of a
house (less freight and lot) on a regional or national basis. Further

112

than this, they will have cut down to a minimum the paper work
and costs involved in selling, so that title search fee, insurance,
amortization, interest, possibly taxes, and even maintenance payments
are all included in the only two figures which the homebuyer will
have to consider: down payment and monthly payment.

It may also be expected that the dealer, if he is to maintain a high
sales level, will increasingly become a guarantor of performance of
the product and an expert service man. One-year guarantees are
already given in many cases. This will become almost universal, as
will a high level of servicing of all sorts, possibly as part of the
purchase price.

6. Simplified Financing

Unquestionably the emergence of well-advertised brand-name
houses, in combination with a continued or expanded program of
government mortgage insurance, will tend to turn the mortgage into
a more negotiable form of earning asset. This will fit in with the
growing tendency for families to purchase houses out of current in
come rather than savings. It is possible that the trend will be in the
direction of forms of tenure and home financing which combine
ownership and tenancy in some manner, as, for example, the pur
chase-option plan. Prefabricators may be the first to introduce such
a scheme on a wide basis.

The nature of interim financing (short-term or construction financ
ing) may be expected to alter as the house is increasingly industrial
ized. A common future procedure will be the combination of chat
tel and real estate mortgage financing in which a finance company
will pay the prefabricator for his package at the time of shipment,
advance funds to the dealer for site improvement and erection and
completion of the house, and sell the final mortgage to portfolio
investors.5 In this way the final mortgage lending institution does
not enter the picture until the completion and sale of the house, and
interim financing is secured less and less by the house itself and
more and more by the general assets of the growing prefabrication
enterprise.

5 This scheme was suggested to PHMI by John Richardson in 1948.

113

7. Sales Cost

Regarding the sales aspects of marketing, there is bound to be a
growing realization of the importance of effective advertising and
sound sales techniques— involving greater expenditures than are gen
erally allowed for at the present time. The industry has been re
minded that it pays for advertising on the average only about $1
per $7,000 house as compared with the automobile manufacturer's
average expenditure of about $10-$15 per $1,500-$2,500 automobile.
Yet insufficient allowance for the cost of selling at the producer's level
has been the admitted cause for failure of more than one promising
company.

8. Sales to the Government

As the participation of government in housing increases there will
also be increasing opportunities for group sales to agencies of the
government, and special attention will have to be paid to this sort of
business since it is inherently different from regular private business
and has rules of its own. Among these is an old maxim: in selling
to the public, sales costs count; in selling to the government, pro
duction costs count.

9. Sales to Operative Builders

Sales to large operative builders whose projects are generally
identified only with their own names may become an important part
of the total prefabrication business, and a few companies may con
tinue to make this a basis of their pattern of operations, contenting
themselves to carry on a sort of anonymous refining stage in the
housebuilding process and organizing their plant facilities for a fluc
tuating volume of production. Nevertheless, such sales will tend in
the future to be more interesting to the manufacturer of house com
ponents than to the prefabricator of finished or nearly finished houses.
For in addition to causing uneven production, the large orders of
operative builders usually involve little chance for disclosure of the
manufacturer or trade name to the final purchaser, and great pres
sure for variation in the product in view of the size of the order.

114

The manufacturer finds it hard to build up in this way the all-out
mass advertising, sales, and distribution required for mass produc
tion. Further, the operative builder, with little fixed investment com
pared to the prefabricator, can remain inactive when things are bad,
and return to compete independently ( as he has in the last few years )
during a seller's market. This indicates a potential advantage to the
manufacturer of components which the manufacturer of trade-name
houses will seek to overcome by working with site developers who
find value in his trade name, by building up his own site-develop
ment teams, and, at the same time, by diversifying his sales as fully
as possible. Such men feel strongly that, even from the point of
view of a potential investor, the surest protection lies in the diversi
fication of risk and in carrying the advertising and trade-name rela
tionship right through to the ultimate purchaser, as has always been
done in other mass-production industries.

10. Market Analysis

The attempt of prefabricates to get a sound market analysis
is complicated by the nature of the housing market in general.
There is a growing realization of the fact, pointed out by William K.
Wittausch,6 that competition between prefabricated and conven
tional houses is overshadowed by competition between any kind of
new house and the supply of existing houses. As times become bad,
the owner of an old house can sell it at less than the production price
of a new one; and, generally speaking, an old house in a good loca
tion will sell better than a new one in a poor location, particularly
if the new one is also very small. There is no assurance that there
will continue indefinitely in the future to be a ready market for the
mass-produced minimum standard house; some signs indicate that
the stable market in the future may be rather for houses featuring
good value at low rather than minimum cost. The private indus
trialist will come to recognize that the purchaser of even the lowest-
price mass-produced house is in all probability making the largest
single purchase of his lifetime, and he will not be tempted solely
because a house is in fact and in advertising claim a stripped-down,
rock-bottom minimum.

6 "Marketing Prefabricated Houses/* Harvard Business Review, XXVI (No
vember 1948), 696.

115

11. Seasonally

Another aspect of the housing market which bears attention is its
seasonal variation. Many a prefabricator seems to assume that this
will be eliminated when the production operations are handled in
the factory and the remaining local work is efficiently scheduled in
advance. It should not be forgotten, however, that seasonality in the
housing market is in large measure a reflection of seasonal forces
in the lives of the families concerned. School terms, spring clean
ings, June weddings, and summer vacations will continue to be im
portant factors after production and erection have been put on a
twelve-month basis. Although it doubtless can be reduced, seasonal
variation may never be eliminated; in all probability it will continue
to have an influence on costs and prices.

12. The Special Nature of a House

The largest marketing problem is found in the fact that houses
are not mere consumer goods, to be used and thrown away when
they fall apart. They are the focus of the basic social unit in our
society and a natural locus for complex social drives and taboos,
for unreasoned preferences and idiosyncrasies. Prefabricators are
finding that it requires far more skill to mass-produce and market
than is generally recognized. This is not understood, certainly, by
most of those who would have us believe that the housing industry
is completely out-dated and ridiculous. Something like a ball-point
pen or a television set, designed to satisfy a relatively new, special
ized, and uncomplicated demand, may be manufactured on a fairly
logical basis and sold with relative simplicity; prefabricators deal
with a real problem in marketing the family home.

III. Future Problems within the Industry

The major future problems arising out of the industry as we know
it today include few that are new or unexpected. Yet their very sim-

116

plicity and obviousness have tended to make them easy to forget,
and for that reason they are briefly summarized here.

A. Central or Branch Plants

Prefabricators will frequently have to decide whether to expand
central plants or open branches. For houses manufactured of rela
tively conventional materials, the problem may be solved indirectly,
as a result of combinations and integrations which bring a number
of separate plants into one large procurement, production, and mar
keting combination, with production or assembly remaining local
ized. For the metal houses, a large central plant may be more logi
cal, although component parts could be made by a number of large
manufacturing plants and assembled at localized assembly plants;
both types of operation have been attempted and so far the choice
between them is not clear. The temptation to set up a large, efficient-
looking, central production plant will be strong, and such a plant
will have no small value as a device for giving both the public and
financial circles a tangible spectacle of efficiency, large assets, and
stability. For houses of concrete and comparable bulk materials,
there seems little likelihood of the development of centralized pro
duction or even assembly plants, excepting for specialized compo
nents or in areas of unusual concentration of demand. More likely
will be the continued development of mobile or portable production
machinery and equipment, designed to be set up at the site and to
effect great economies when a large number of similar units can be
produced within a short radius of operations. Such equipment is
particularly suited to the construction at one time of an entirely new
project or community under a single developer.

B. Site or Factory Fabrication

It is sometimes argued that good site fabrication makes prefabrica-
tion unnecessary, and certainly site preparation and fabrication tech
niques will develop hand in hand with production techniques gen
erally and will help to effect a general reduction in costs. As the
prefabrication plants become increasingly efficient, and as substan-

117

tial cost savings in the house package become available, however, it
is reasonable to believe that even the site developers will find it
advantageous to purchase many of their units from the prefabri-
cators. Site developers may also turn to the prefabricators because
of the advantages of shifting to them the worries about procurement
and delivery and because of the possibility of cutting thereby the
time and expense required for construction and financing. At the
same time, as we have seen above, the dealers in prefabricated
houses will become more interested in large site development. The
problem thus becomes not one of choice, but one of taking fullest
advantage of both fabrication methods.

C. Low Price or High Value

Although low-cost houses will necessarily continue to be the major
market for prefabricators, the industry generally will be faced with
the problem of deciding when the time has come to seek better values,
by adding space, equipment, or facilities, instead of lower prices.
For houses purchased with government aid, there seems every reason
to believe that space standards will be moved up from the minima
which have prevailed during the last few years; and when the house
is privately purchased, owners may become increasingly conscious
of the illusory quality of a bargain purchase which proves to be
unsatisfactory for normal family living. Space can be added more
easily and cheaply than many other features of a house, and, as its
great value is understood, it will be increasingly demanded in the
future. At the present time, however, space in the house is very
hard to merchandise.

The manufacturers who are most interested in better values are
those who have found that they can market their product more easily
if it contains certain special features or pieces of equipment which
on the normal market might come under the heading of luxuries.
Mass-produced as parts of the house, such special features probably
add little to the cost but a great deal to the salability. Obviously,
this can be overdone. The ingenious prefabricator will be careful
to develop and include just what is necessary to give his house a
special appeal at the best price possible. For a while this kind of
gadgetry may have the effect of reducing the real quality of the
house in the interest of including more sales features. There is rea
son to believe, however, that competition within a highly indus-

118

trialized housing industry and the examples set by the government
in its programs for the lower-income groups, together with a growth
in the store of general knowledge regarding basic physiological and
psychological requirements, will counteract that tendency.

D. Evolution or Revolution

Of the general problems troubling those interested in prefabrica-
tion, one of the most interesting is the problem of evolution versus
revolution. Evolution is a normal, familiar process, and many argue
that it alone can succeed. The argument for revolution may be
summarized as follows: factory methods do not promise a reduction
in wood-processing costs sufficient to offset the increases in over
head expense which result from moving the operations from the
site to the factory; therefore there is little hope of developing a
genuinely low-cost home through evolution along conventional lines,
and hope must be placed in revolutionary production techniques,
probably making use of metal as a basic material.

Against this must be weighed the difficulties of creating an entirely
new production process and the necessary accompanying marketing
process.7 It is generally easier to create an entirely new industrial
operation to produce new articles, whereas for the mere improve
ment of old articles, the obvious pattern is a development of old
industries. The house-trailer industry in this country sprang up to
produce a new article, which offered a service not fully performed
by any existing product (although it could also be used as a house
when fixed in place); despite its very limited probable market, the
industry grew up very rapidly, and recently has been selling more
than twice as many units as the prefabricated housing industry.8
Perhaps this is a trend towards a new way of living and a step in the
revolutionary process. Where the attempt has been made to set up
an entirely new industrial organization to produce houses in the past,
money has always run short before competition with the existing

7 John Ely Burchard has presented a good discussion of this point in "Pre
fabricated Housing and Its Marketing Problems," The American Marketing
Journal, II (July 1935), 150-6.

8 In 1945 the trailer industry produced 16,225 units; in 1946 it shipped 47,103;
and in 1947, 70,078; with a further increase in 1948. Figures from Facts for
Industry, Series M45A-68, U. S. Department of Commerce (October 11, 1948),
Table 1.

119

industry could be effective. "The sun's rays of capital have been
applied often intensely but never for long because no one could
afford a sustained effort/'9 Furthermore, there are technical diffi
culties to be overcome in producing a house equally suited to cli
matic conditions in Minnesota and in California, and careful design
is required if the product is to be both mass produced and non-
uniform. And those operating at tremendous scale from the start
have little basis in practical experience for finding a realistic com
promise between a highly functional product which might not sell
well because it is unpopular and a more conventional product which
might not show enough saving in cost.

The likelihood is that the natural process of evolution and the
earnest attempts at revolution will both continue in the future, and
that what might seem in prospect to be revolutionary will seem in
retrospect to have been merely evolutionary.

E. One Model or Many

Another problem arises in deciding whether to produce a single,
or at the most a very few, standard models offering the best plans
possible for average buyers and a consequent efficiency in both pro
duction and marketing, or to make a line of component parts which
may be assembled to suit individual tastes to a much greater degree.
There may also be intermediate stages between these extremes; for
example, a company selling a line of many models which are as
sembled by varying the numbers and arrangement of a relatively few
standardized partial assemblies or components. A certain amount
of variety in the product is compatible with mass production, as can
be illustrated by the automobile industry, particularly by the recent
lines of automobiles in which standard component assemblies may be
interchanged not only in different models of a single make but even
in different makes of cars, from the smallest to the largest. Further
more, we have seen that it is possible within the requirements of
mass production to make allowance for a certain degree of regional
difference through partial assembly by the distributor. It seems
likely that the lowest costs will be achieved when the product is
the most fully standardized— if a mass market for so standardized a
product is developed. If the mass market cannot be fully developed

*Burchard, "Prefabricated Housing and Its Marketing Problems," p. 152.

120

for a highly standardized product, the line of models made up from
standardized assemblies should prove to be the most economical.
Undoubtedly, there will also develop a substantial market for simpler
components such as panels, manufactured for general distribution
as a sort of superior building material and for assembly by the local
builder or architect. Unless a large project is being developed at one
time, however, this pattern of operations will almost surely lead to
somewhat higher cost to the ultimate consumer in return for in
creased individuality. Finally, even those who prefer and can afford
to have their houses built individually for them will take increasing
advantage of the availability of manufactured assemblies and com
ponents.

F. Optimum Level of Standardization

The previous question tends to become one of the optimum level
of standardization: whether at the 4" building material module, the
modular panel, the three-dimensional section, or the completed house.
Undoubtedly all will be under development at the same time, and
ultimately all may be the basis for a true mass production. The
4" module can be assumed to be already well on the way towards
this goal, and it might be argued in any case that differences in the
character and purpose of its development rule it out of this discus
sion. These are differences, however, only in degree.

G. Duplication by the Conventional Builder

Overriding all these problems, from the point of view of the pre-
fabricator as we now know him, is the problem of the ease of dupli
cation and the adoption of his new techniques by the conventional
builder. The conventional builder has been criticized by many as
old fashioned and unlikely to compete along the paths of industriali
zation. Actually, as we have seen, many industrial techniques are
already turned to his use as well as to that of the prefabricator.
Aluminum siding and roofing are widely marketed. Even the highly
industrialized vitreous enamel finish can be purchased for home use
from manufacturers. Such vitreous enamel sheets are thin enough

121

to have many of the characteristics of wallpaper. It is this quick
utilization by others of his developments which illustrates that the
prefabricator may serve primarily as an agency for the first substan
tial penetration into the building industry of modern mass-production
theories.

IV. Larger Housing Issues

Lord Kelvin has said that one measurement is worth a thousand
questions. In the prefabrication field, as in many others, this is not
always true. With relation to some of the most important forces
bearing upon the future of housing, even the basic theories have
hardly been developed. For those dealing with such forces, one
question may be worth a thousand measurements. It is the purpose
of this section to raise some of these questions. These are often
easy questions to ask; unfortunately, for most of them there is little
indication of a satisfactory answer.

A. The House Itself

1. Shrinkage

The house in which the average family lives has been undergoing
a steady change in character in recent years. More and more of
the functions which used to be performed within its walls have been
transferred elsewhere, while in some degree there has been a replace
ment by functions not previously considered part of the house. Thus
food preservation and preparation require a very small portion of
the time, energy, and space formerly devoted to such activities. The
recent introduction of the home freezer and of other specialized
kitchen equipment represents not so much a reversal of this trend
as the provision of new types of conveniences. Rooms for formal
entertaining and space for making and washing clothes have been

722

curtailed sharply, and, here again, the recent growth in popularity
of home entertainment devices and sewing and washing machinery
takes the form of an added service for those to whom the commercial
facilities were unsatisfactory. Space needed for heating equipment
and fuel has been sharply reduced; servant quarters are fast disap
pearing; and rooms themselves are becoming smaller. All this has
been reflected in a contraction in the size of the "average" house
over the course of the last few decades.

In the future, to what degree will this shrinkage continue? Are
there practical limits to the reduction of meal-preparation and eating
space, or will there be a further contraction, possibly with the devel
opment of precooked and fresh-frozen full-course meals, specialized
catering services, and the like? Will families go out for more and
more of their entertainment, using the automobile to get them to
commercial and community recreation centers, or will the radio and
television bring them increasingly back into the home? Large formal
occasions tend already to be celebrated out of the home in rented
quarters. Will this be true for smaller occasions also, as better social
rooms become available on a neighborhood basis?

This list of questions could be expanded, but it is enough to illus
trate the point: an increasing reliance on the community means not
only a shrinkage in the size of the house, but also an increasing
community influence upon the family enjoyment of the house. The
successful prefabricator will be prepared not only to modify his
product, but also to pay increasing attention in selling, locating, and
erecting his houses to the character of services now being performed
by the community.

2. Mechanical Independence

Will there be an increasing trend towards mechanical independ
ence? We know that, although families are steadily moving into
the large metropolitan areas in this country, within these areas they
are moving rapidly out from the centers to the suburbs. This subur
ban movement has been speeded by the automobile and by the
availability of electric power, and a boost was given by such me
chanical appliances as the washing machine and the home freezer.
Will there be a further development along these lines with further
decentralization, or will peacetime living bring a return to the pre
war inclination to purchase services from specialized and centralized

123

organizations? For example, if it were possible to have electric
power as cheaply from a home generator as from public service com
panies, would the average family wish to have this added independ
ence? An efficient and economical chemical toilet, with or without
the re-use of the water, could make significant changes in the con
struction of the house and even in the structure of the sewer-bound
society in which we live.

For the average person an increased provision of home machinery
and equipment might be attractive, but it would substantially in
crease the first cost of the house, and this the average family cannot
stand. Although in some respects it is not in the best interests of
those concerned, the trend towards designing for low initial cost will
very likely continue, even in the face of higher maintenance and
service expenses. This trend may be reversed principally in the
construction of mass housing, whether supported by government or
built as an equity investment; in either case there are clear advan
tages in paying a high first cost which will be more than balanced by
long-range efficiency.

3. Flexibility

From another point of view, it is important to know the degree
to which the house must be made flexible— to permit changes in
size and arrangement with the changes in the composition and char
acter of the family living in it. Many a prefabricator has been per
suaded of the need for such flexibility, and, because his construction
system offers easy demountability, this tends to be made a selling
point. A large segment of the public, certainly, has expressed the
desire to add bedrooms, shift plans, and generally have an "expansible
house." Yet, it is fair to ask, to what degree is this desire real and
to what degree imaginary? How many average householders have
carried out extensive remodeling of their houses in recent years?
How many more would have done so were it inexpensive and easy
to do so? The frequency of family moving in this country may be
enough to take care of such adjustments. Family pattern is partly
a matter of size and facilities within the house and partly one of
location in a general sense (urban, suburban, or rural— to obtain cer
tain definite benefits, real or imagined), of a desire for gain in social
status, and of a complex of other factors. If there is no trend to-

124

wards a fixed location for the family, there may be no particular
need for great flexibility in the house.

Where it can be easily provided, however, as by movable parti
tions, little is lost in terms of cost and much is gained in sales appeal
by providing flexibility. Further, if stable communities should de
velop having a good cross section of types and sizes of families and
serving them well enough to reduce the urge to move, then there
might well be advantages for the prefabricator offering flexibility
in the form of standardized building components for individual
assembly and easy reassembly according to need.10 On the other
hand, several prefabricators propose to offer frequent new and im
proved models and to persuade the homeowner to trade in his pres
ent house for the latest model. This would offer another sort of
flexibility, if it were easy to detach houses from the land and trade
them about like chattels, or if land planning trends and social devel
opments in the future should make it less of a problem of adjust
ment for a family to find a new location for each new house. If
models were traded in only at the time of major family changes
or moves, of course, the problem would tend to take care of itself,
since these changes often are accompanied by changes in location
preference within the community. For example, young married
couples like to live in central locations, while parents of small chil
dren prefer open development, freedom from traffic, and suburban
informality.

4. Single-Family or Multifamily Units

Another important question is this: will the mass-produced units
of the future be single-family or multifamily houses? It is often
claimed that multifamily homes can be offered at slightly less cost,
and that, with the benefits of the best of modern design, they offer
certain advantages to the family. In fact, when it comes to very

10 Suppose a family, growing through the years, has reached the stage of
adding a "Cadillac grade"' mechanical core and a great deal of living space
enclosed by standardized wall panels. Then, when the daughter marries and
moves away, it becomes possible quite literally to break up the old home, giving
her the old "Ford grade" core and enough panels for a small house in which
to begin married life. This kind of speculation tends to minimize the prob
lems of foundations, gardens, and land use in general, but it has a certain fasci
nation, nevertheless.

125

small units occupying narrow lots of land, many feel that multifamily
units are definitely superior, offering better space with increased
privacy. High apartments also have their strong advocates. Yet
the prefabrication industry has attempted very little as yet along
multifamily lines. This situation will be altered in the near future,
no doubt; construction systems will increasingly be made adaptable
to multifamily structures; and for the high fireproof structures, special
systems will be worked out to take advantage of components adapted
also to simpler construction. These developments will be accom
panied, and greatly abetted, by two other developments: the growth
of modular coordination and the increase of low-cost project-type
housing built by public agencies, by large developers aided by the
government, and by equity investors such as the large insurance
companies.

5. Durability

What is the optimum durability of the house? Prefabricated houses
have in the past suffered from a popular belief that they were "tem
porary" houses, when the fact is that the industry might better won
der whether it has not been building too well. Two arguments are
often put forward in favor of decreasing the length of life of the
average house. The first is that long life means rigidity, whereas
family requirements change, land use patterns change, and our whole
way of life changes; in short everything changes except the house
in which life is supposed to take place, and that is altered only by
the addition of mechanical equipment and conveniences and by
minor adjustments in the details. The second argument is that if
houses were less durable, more would have to be replaced each
year; the building industry would have a larger constant core of
replacement building; larger-volume production would in turn lead
to more efficient production; and fluctuations in building activity
would be less extreme. The first argument can be answered in part
by flexibility and good planning, and the second may be challenged
on the basis of cost and practicality. Can the nation afford to
replace housing on the basis of a life span of definite and rather short
length? The advocates of greater replacement might ask whether
it can afford not to do so. However, the building of a house calcu
lated to last an exact number of years is no mean feat, and experience
with temporary structures in the past has shown that the life of a

126

house depends more on maintenance standards than on construction
standards. Enforced replacement is hardly an immediate prospect.

6. Obsolescence

To this must be added another question, that of obsolescence.
Whether we like it or not, the fact is that most of our large-production
industries depend in part upon a rapid rate of obsolescence. This
may be natural for some products, but it is largely artificial when it
comes to radios or automobiles. Actual or pretended improvement
in performance leads to a demand for the new product while the
usefulness of the old one continues largely unimpaired. Something
of this sort is very likely to appear in the prefabrication industry in
the future, and it may have an important bearing on the question
of durability as well. Would one build an automobile to last 60
years if constant or significant technical improvements were antici
pated? But obsolescence will not become a major force unless there
is also developed a second-hand market. A house goes into use
only after it has become a piece of real estate, attached to a certain
piece of land. A second-hand market in house packages would re
quire that dealers become real estate operators on a large scale in
order to put the houses quickly into use, and that land use and home-
owning customs undergo a sort of revolution. This is not impos
sible, of course, but it does not seem likely to come to pass in the
near future, at least in anything like this form. It seems far more
likely that any second-hand market which develops will follow the
lines of the traditional real estate market,11 and that there will be
little selling of used houses without lots for some time to come.
The costs of moving houses and making and breaking utility con
nections are too great, although we have seen that new developments
may one day take care of even such problems as these.

11 It may be noted in this connection that the new president of Gunnison Homes,
Inc., is a man with vast experience in handling real estate.

727

B. The Community

1. General Problems

In addition to an understanding of the changing character of the
house itself, the prefabricator must have an appreciation of the extent
to which it is dependent on external factors for satisfactory perform
ance. Two quotations from experts will serve to give an indication
of the factors involved. William J. Levitt, the well-known Long
Island builder, points out:

There is no such thing as a complete, factory-engineered house— because no
one has discovered how to prefabricate the land, how to prefabricate the
road in front of the land or the water main that goes into the house.12

Russell W. Davenport, moderator of the Life Round Table on
Housing, concludes that most of the trouble with prefabrication lies
in the nature of the product itself.

A house, in short, is not merely a mechanical product; it is not even
merely a physical or material product— though even on this plane stand
ardization and mass production are difficult. A house transcends the physi
cal and transcends the tangible to become part of its surrounding civiliza
tion. It is a civic or social product; and for those who live in it it has a
spiritual significance. These elementary facts must constantly be borne
in mind if our efforts to house ourselves better are not to meet with dis
aster.13

Many aspects of his market are beyond the control of the prefabri
cator, no matter how large he may be, and can be influenced only
by public understanding and action— for example, a boom in specula
tive land prices, a series of municipal "protective regulations" which
in effect require excessive development costs, or a blight of excessive
land subdivision and clouded land titles. The prefabricator should,
however, make his plans and conduct his operations with an intel
ligent regard for these broad problems.

™Life, 26 (January 31, 1949), 74.
« Ibid., 78.

128

2. Future Demand

One important element of his plans is the estimation of housing
demand in the future. This is the most complex sort of problem,
involving as it does everything from consumer tastes to government
policies, and yet market-analysis techniques and data are woefully
inadequate. Important considerations are the supply of existing
houses and the measures provided for the demolition of those houses
which are obsolete. The industrialist may argue that disposal of the
obsolete will follow naturally from an abundant production of the
new. If the obsolete drops sufficiently in price, however, and is
usable, can the new reach abundant production? How does this
take account, moreover, of the investment in developed land, utili
ties, and community services? How long can we afford to concen
trate new development on the outskirts of our cities and let blight
move in behind? These may be the problems of the city planner
and the investor in real estate, but they are also the problems of any
mass producer of houses.

3. Competition from Existing Houses

The housing market displays the characteristic of rapid obsoles
cence at the top and very slow obsolescence at the bottom. When
houses have become actually unsuited to human habitation, energetic
exercise of the police power will suffice to tear them down. Above
this level, however, they remain a problem and a source of competi
tion for any type of new housing. As a result, the prefabricator may
decide to operate at higher price levels, counting on producing new
models attractive enough to entice former purchasers to trade in their
old houses. In this way the old houses are to be started on the
"filtering down" process by which low-price housing becomes avail
able at second, third, or fourth hand to those who cannot afford to
purchase new houses. In the automobile field, this process works,
and one can purchase a car for $200 whicn is far better than a new
one built to sell at that price. In the past, however, the houses
which have filtered down in this way have been too few and too
poorly adapted to the need of those in the lower-income brackets.
Much will depend in the future upon reaching a low price level for
new houses, so that they may have a broad market from the start.
Otherwise, prefabricators hoping to serve the whole range of hous-

129

ing needs will have to trade purchasers out of their new houses at
an impossibly high rate in order to start secondhand houses down
the line in sufficiently high volume. Large-volume production might
thus require such a combination of low price and high sales appeal
that the old house will be traded in as often as every 10 years; other
wise the prefabricator may become the victim of the housebuilding
cycle as it has operated in the past.

4. Problems of Turnover

There are many difficulties in obtaining a rapid turnover of this
sort. One is the likelihood that the new features upon which the
manufacturer must rely for sales appeal will tend to be mechanical
equipment and gadgetry which may relatively easily be purchased
and installed in the old house, the shell of which is likely to dete
riorate at a slow rate by comparison with its equipment if the level
of maintenance is good. Another is the fact, which cannot too often
be mentioned, that houses are attached to the land. Unless houses
can be made demountable and sold through secondhand dealers like
automobiles,14 the purchase of each new house means moving to a
new site, and usually a new neighborhood. What does this imply
for the stability of communities, for the interest of people in the
local governments and schools, and for those— especially children—
who find adjustment to new social circles a personal strain? What
happens, as a practical matter, to the well-kept lawn and the garden?
If we avoid providing for expansibility or demountability because we
expect our current high degree of family mobility to be maintained,
then we must provide correctives for the problems which such
mobility creates.

5. Community Planning

Broad problems facing the prefabricator often stem from problems
of community planning. The rapid development of a large outlying
tract with hundreds of similar small houses and insufficient com
munity services and amenities, which appears to be the most eco-

14 For an examination of this idea, see Neal MacGeihan, "The Myth of the
Low Cost House," Prefabricated Homes, January-February 1945.

130

nomical manner of providing houses in terms of first cost, may in the
long run so prejudice the housing market that the effects will be felt
by the prefabricator himself. He must seriously consider whether
this sort of entirely unofficial zoning into a one-class, one-income,
undifferentiated community may not be contrary to his own selfish
interests because of the dissatisfaction of those living in such a com
munity—almost certain to be carried over to the house itself. At
present, few have had to worry about these problems, because few
have attempted mass production on such a scale and with such equip
ment and plant that profitable operation over a period of years is
required if the investment is to pay off. In the future, unless the
large producers consider such matters as they grow in stature and
importance, public opinion may compel the local government to take
steps to control them, and they run the risk of becoming in effect
large public utility companies. Through intelligent planning, volume
of sales can be maintained at a high level without injury to the
community from which the houses derive so much of their essen
tial character and quality.15

C. Broad Economic and Policy Problems

Much that might come under this heading has already been
touched upon, but there remain two aspects of the relationship of
the government to prefabrication which deserve consideration here.
Already committed to a public housing program and to a program
of mortgage insurance which leaves the building of many small houses
a matter of private enterprise in name only, the government is taking
an increasing interest in the general field of middle-income housing,
the field of greatest interest to prefabrication.

1. Government Aid

Government aid is not peculiar to housing; it has been widely
used in many fields in the past. The automobile industry, our prime

15 It should be noted that increasing attention is being paid to these matters
of neighborhood planning. The checklist for veterans in For the Home-Buying
Veteran, issued jointly by the several federal housing agencies in 1949, makes
the character of the neighborhood and the character of the lot the very first two
matters of concern.

131

example of mass production, could hardly have developed without
a tremendous subsidy in the form of public roads.16 The day may
come when the government will adopt the often-suggested policy
of establishing a figure below which the production of housing would
not be allowed to fall; under such a policy, if the necessary houses
were not produced by private means, the government would take
over at once. The importance of such policies, and of the govern
ment guarantee of decent housing, will become increasingly large
factors in the future as the prefabricators grow in size and in volume
of production. Self-interest alone should induce the leaders in the
field to take a constructive part in the formulation of broad plans and
to cooperate with the government in setting up that stable market
situation which is necessary for profitable operations.

2. National Capitalism

One aspect of government housing policies which secures a great
deal of attention in business and industrial circles is the emergence
of what has been called national capitalism. In the past, the hous
ing industry has been considered a bulwark of private capitalism,
but there can be no doubt that this is being altered. The problem
is, To what extent? Of particular interest in this connection was
Lustron, organized to produce houses at a scale never before real
ized, and financed initially with $840,000 private equity capital on
the one hand and a $15,500,000 loan from the RFC on the other.
With no further increase in private equity capital, the public loan
later more than doubled, and requests were submitted for increases
to as much as $50,000,000. With a ratio of better than forty to one
of debt to equity, this leverage seemed so great that it was said that
for all practical purposes the government had gone into the hous
ing business.

Many private businessmen were concerned; they believed that Lus
tron had received "favorite son" treatment. Either the government
plans deliberately to take over the housing industry, or it will eventu
ally take it over whether it plans to or not, they argued. Who would
dare to raise the risk capital and create the facilities necessary to
compete with Lustron on an entirely private basis? The government

16 It has been suggested that this theory be carried over into the housing
field, and that the government frankly subsidize housing by the purchase and
free grant of house sites, retaining thereby the control of development.

132

would surely continue its favored treatment in order to protect a
investment. With private capital thus frightened out of the housing
industry, the government would move inevitably towards national
capitalism. Claiming to favor free enterprise, but becoming increas
ingly addicted to close regulation and control, the government might
continue to solicit private capital, but would certainly put up public
capital if none were forthcoming.17 And eventually the same
mechanism would be turned to other fields, said the worried spokes
men of business.

On the other hand, there can be little doubt that many who had
approved governmental support of Lustron were quite free of such
motives. Prefabrication is a young industry, and we have seen that
financing on a tremendous scale is often required for the mass pro
duction and distribution of houses. The risk is so great and the pros
pects of profit so dim in comparison with other investment oppor
tunities that only the government, acting in the broad interests of the
public, can be expected to give such an industrial approach a real
test. When the way had been shown, supporters expected private
corporations to move in quickly and set up competitive enterprises,
and meanwhile Lustron should itself have repaid the RFC and be
come a private industry in the normal sense. Certainly one can
sympathize with the desire to give any likely method of increasing
production and reducing costs a fair chance to prove itself. If
operations should prove extremely profitable, why would large com
panies avoid the field? They have had to deal with the government
before.

17 In his testimony of August 5, 1949, submitted to the House of Representa
tives Banking and Currency Committee, Harry H. Steidle, Manager of the Pre
fabricated Home Manufacturers' Institute, had this to say: "We are therefore
strongly opposed to legislation that would definitely favor any one of several
companies that are heavily indebted to the Government to the disadvantage of
those companies which are in part paying the bill through taxes. . . . This pat
tern of destruction to privately financed producers of prefabricated homes shows
itself in numerous ways, some of which are as follows: (a) in the compulsion
to extend further loans in hope of working out of an already bad situation; (b)
by the practical effect of extending free rent from the War Assets Administra
tion; (c) through intercession before other governmental agencies for the
allocation of steel or other aids not available to privately financed companies; (d)
through authorization of a large sales and public relations staff paid out of
Government loans; (e) by approval of a national advertising campaign paid for
out of Government loans; (f) through pressures of varying degrees exerted on
Government buying agencies to purchase the houses made by the indebted com
pany." The legislation in question, which would have authorized RFC market
ing loans to companies already holding RFC loans, was defeated.

133

It is too soon to know what the eventual result of the intervention
of the government will be. The fully equipped Lustron plant has great
potential value as a producer of houses, of bathtubs and sanitary ware,
and of light structures in general. It contains probably the world's
largest ceramic line. Dismembered and sold at auction, it would bring
the RFC only a few cents on the dollar. Yet the political interest
which has been aroused makes it unattractive to private investors, and
competitors are standing in the wings, fears of national capitalism
notwithstanding. United States Steel Corporation owns a control
ling interest in Gunnison Homes, and its Ambridge research labora
tories are at work on housing problems. Republic Steel Corporation,
through its Truscon division, already supplies a great variety of com
ponents to the housing industry. The aluminum companies have
nearly all come out with lines of building materials. All could move
in fast and make a good fight for the business. Some undoubtedly
prefer to avoid making houses as such and, by manufacturing a line of
highly developed components, plan to take over most of the business
without direct competition.

As a sidelight on this question, it should be pointed out that many
public housers seemed to dislike Lustron with an intensity approach
ing that felt by these businessmen. Conscious of limited objectives
and unintelligent actions on the part of private builders in the past,
they were inclined to dismiss as technocratic pipedreams all efforts
to reduce the capital costs and to increase the supply of housing by
processes of industrialization. While it is true that there has been
some justification for a healthy concern, it seems illogical to be sus
picious of any approach to an increased supply of better housing
which does not involve public agencies and project developments.
Between the suspicions of housers that Lustron was a mere attempt
to discredit the public housing program through a cynical mass-
production mythology and the suspicions of the businessmen that it
was the first step in socializing the industry, the company had plenty of
intangible difficulties to add to its normal production problems.

A final note on these relationships: few business suspicions re
garding the future seem to attach to the work of such government
agencies as the FHA and the VA, through the combined resources
of which it is possible for private builders to put up houses without
the investment of any private equity risk capital whatsoever. Here
the initiative remains in the hands of the local builder, it is true, and
the financing is worked out in local circles, so that the process ap
pears to be more conventional. But at the first major break in prices
and employment, the government will take over a large share of our

134

housing supply. Clearly, the government program needs to be con
sidered as a whole.

V. Conclusion

A number of questions have been raised in this chapter, but the
problems extend beyond any single set of questions. Above all it
should be clear that the prefabrication industry faces problems of
very great variety, many of them far more complex than is generally
recognized. To analyze these problems and to work out the means
of finding significant answers is the job of research, and no one can
doubt that a great deal of research is needed. By and large, the
technical questions, while easier to answer, tend to seem almost un
important by comparison with broad questions of economics and
sociology. And yet even technical questions often require great skill
and patience. For research is always a long process— in building
especially so— and it is even longer before practical application takes
place.

Those familiar with the state of knowledge and research at any
time can make fairly accurate predictions regarding the develop
ments likely to occur over the next period of years, subject only to
accelerations and decelerations resulting from such factors as wars
and depressions. Everyone is aware of the detail in which such
imaginative writers as Jules Verne and H. G. Wells were often able
to forecast events which have since transpired. Ordinarily, there is a
substantial time lag between the day when knowledge justifies a
prediction and the day when the prediction comes true. The exist
ence of this time lag makes it possible in normal times to foretell
whether or not magical industrial advances are likely to take place
in a given field in the next few years; there is little in the current
state of building knowledge and research which suggests that any
such advances may soon be expected. It is largely this belief that
we were in a period of comparative calm which served to justify
us in exploring in so great detail the existing state of the industry.

For such new ideas as may now have reached the stage of clear
anticipation, there remain long periods of development to be under-

135

taken, first in the laboratory, and later at pilot plants. Such is the
skill of modern engineering, however, that this process could be
carried out easily, provided enough money and energy were put
behind the development. Yet, with technical development com
pleted, the new idea must buck other forces which oppose the in
troduction of any innovation— the so-called barriers to technological
advance which have frequently been described.18 For even the
simplest ideas, therefore, a widespread application may be long
delayed.

Most of the ideas in prefabrication, furthermore, are not simple.
They involve questions many of which could be answered compara
tively easily if they could be put in direct, technical form; the
trouble is that in this field few questions can be put in that form.
To illustrate this point, take the question of corrosion. Prefabricators
with corrosion problems may draw on scientific facts which have
been well established for years, but they are more concerned with
satisfactory performance at low cost than with scientific advances.
The basic research in corrosion is therefore primarily conducted in
laboratories little concerned with the problems of housing.

This illustration might be multiplied, but it will suffice to point
up the fact that little scientific satisfaction is available in the field
of housing, where every problem is confused by considerations of
economics, sociology, physiology, and psychology. The result is
that, in the universities and elsewhere, research men have preferred
less complex and more satisfying problems.

During the war it was possible to attract to government war
research a great number of the best scientific minds in the country,
despite the fact that their work, with very few exceptions, was not
scientific research at all, but rather the accelerated development, for
war purposes, of scientific knowledge derived from research done
as much as two generations earlier. Unquestionably there is an
emergency in housing today, but the ^ense of urgency and of over
all organization along lines of clear and definite policy has been
missing. Such research as is being done frequently represents a
search for suitable compromises limited by the special interests of
sponsors, by lack of resources, and by the absence of programs broad
enough to challenge assumptions and seek far afield for determin
ing forces.

is See, for instance, Bernhard J. Stern, "Resistances to the Adoption of
Technological Innovations," Technological Trends and National Policy, National
Resources Committee (June 1937), pp. 39-66.

136

We have pointed out that when an entirely new product is devel
oped a new industry will often be created. But houses are not new
products, and they cannot quickly be "rationalized." Men may select
a radio with a relatively dispassionate logic, but emotions and tra
ditions tend to dominate in the choice of a house. Obviously, care
ful sociological research is needed even to identify the main drives
operating in this field, and much more research will be needed before
we know how to direct these drives, or to what ends. At the present
time, the beginnings of sociological studies in housing have been
made. Of especial interest are recent studies made by the Research
Center for Group Dynamics, of the University of Michigan,19 and by
Robert K. Merton at Columbia.20 The work has barely been begun,
however.

Far more is involved than the tabulation of preferences regarding
the size and arrangement of rooms. It may one day be shown, for
instance, that satisfaction with a house depends less on the character
of the house itself than on the social relationships formed by the
family. The market may grow in the future for well-planned projects
of small houses balanced by good neighborhood facilities. Even
such broad considerations as full employment and increased leisure
will have their influence on the product and on the industry.

Lacking basic and fundamental facts in all these situations, we
may seek empirical data on which to base decisions in the imme
diate future through a careful analysis of the activities of such pro
ducers as Lustron, such builders as Levitt and Sons, and such gov
ernment activities as those of the FHA and the FPHA. In the past,
following a national tradition of never looking back, we have been
guilty of shocking waste through our failure to profit from the great
experiments and projects we have built. We can no longer afford
such extravagance.

H. G .Wells pointed out that the rapid rise of the Germans in na
tional strength and importance in the nineteenth century could be
attributed in large part to their discovery that knowledge was a crop
like any other, to be increased in quality and in yield by cultivation
and by the intelligent use of fertilizers. This lesson the recent war

19 Their original work in this field, sponsored by the Bemis Foundation, is de
scribed in the book by Leon Festinger, Stanley Schachter, and Kurt Back, Social
Pressures in Informal Groups: A Study of Human Factors in Housing (New York:
Harper, 1950).

20 Reference to many other studies may be found in "Selected References on
Family Living Requirements and Public Acceptance Factors Relating to Hous
ing Design," HHFA Technical Bulletin, no. 4 (April 1947).

137

has finally brought home to us in this country; let us apply it in the
field of housing.

One may ask who should do the research work: private companies,
industrial associations, educational institutions, professional societies,
or government agencies? Obviously, the answer is: all of these, in a
coordinated effort.

All now are needed, and the way seems open at last for all to take
part. Private producers are becoming large enough to devote serious
efforts to research; industrial associations are growing in importance;
educational institutions have increasingly entered the field; profes
sional societies have taken initiative indirectly and directly in the
stimulation of new research, as typified by the formation in the Na
tional Research Council of the Building Research Advisory Board;
and the government is now provided with legislative authority for
a large-scale program of research in the HHFA.

The importance of prefabrication in helping to stimulate this re
search effort lies in the fact that, because of the problems inherent
in adopting and executing a suitable pattern of operations covering
every step from the procurement of raw materials to the servicing of
the final houses, it has brought sharply into focus the needs for re
search, the possibilities and difficulties of industrialization, and the
special complications of the production aspects of the housing prob
lem. To return to the thought expressed in the introduction of this
chapter, it may well prove in the end that prefabrication has been
only a local and specialized advance within a broad process of in
dustrialization, and that in the future there will be little point in
trying to decide whether or not a housing process can properly be
called prefabrication. The prefabrication industry has served, how
ever, as an almost ideal framework in which to study the overall
problems of housing.

118

Part A A •

Chapter

INTRODUCTION

This part of the book is devoted to a detailed and, as nearly as pos
sible, factual and objective analysis of 130 of the prefabricators whose
production facilities were visited and representatives of whose man
agement were interviewed during the course of an extended field
survey.1

No one survey could give the definitive story of prefabrication as
a whole; yet it has been possible to describe in some detail the
activities of a large and entirely representative portion of the in
dustry. All but a very few of the leading companies are included
in the 130 analyzed, and a particular effort was made to include
companies promising the greatest innovations, whether or not they
were in actual production at the time.

In this analysis, the various methods, designs, and facilities are
discussed primarily in terms of the number of companies making use
of them. This was a necessary procedure, for accurate information
on production was often not available, and in many cases the value
of an idea could not be fairly judged by production figures. Analysis
by numbers of companies also has its weakness, however. If Lustron
had reached its expected rate of production, for example, it would be
making more houses per year than have been sold by the entire in
dustry in any single year in its history. From the viewpoint of the
general housing market, therefore, a decision by this one company
might have importance far beyond the apparent meaning of our
figures. On the other hand, our principal interest in this part of the
book is in finding out what patterns of operation were being used at
the time when the greatest number of companies was active in the
field. This sort of information can best be approached by the method
which we have adopted.

The discussion is broken down into a consideration of five basic
components of a pattern of operations:

Management

Design

Procurement

Production

Marketing

and the treatment is factual wherever possible. Factual treatment is
not always possible, however; for example, the prefabricators'

1 Material regarding the methods used in this survey, lists giving full names
and addresses of companies visited, and other reference data are included in the
Appendices.

141

thoughts regarding the government or labor can be reported only as
opinion, although it is opinion based upon interviews, press state
ments, and actions they have taken.

The bulk of the material in this part of the book was gathered
during the Bemis Foundation's field survey in 1946 and 1947, but
references to more recent developments have been included when
these would help to give a full understanding of the problems in
volved or of the trends within the industry today.

Because of the organization scheme which has been followed, there
is in this part of the book some duplication of material presented in
the first part. There, the references were usually brief, however,
and they served primarily to illustrate general points under discus
sion. Here, interest is centered on specific details of the prefabrica-
tion process.

142

Part A X •

Chapter

MANAGEMENT

I. Background

Many of the differences in patterns of operations of prefabricators
may be attributed to differences in background, that is, in the nature
of the business from which the prefabrication business developed
and in the previous experience of top management. Unquestionably
many costly mistakes have been made by carrying over to this new
industry techniques which were more familiar than suitable. On
the other hand, background can explain the success of certain com
panies in dealing with the very specialized conditions of a local
market. In the companies analyzed, the following types of back
ground were noted:

Frequency
Previous Experience (per cent)

1 Building contractors, construction engineers, and operative

builders 31.2

2 Building materials manufacturers or salesmen 19.5

3 Architects 19.5

4 General manufacturers (including shipbuilding, boxmaking,

light metal fabrication, and heavy industry) 17.5

5 Salesmen 6.5

6 Bankers 2.6

7 Lawyers 1 . 9

8 Other 1.3

Without exploring in detail the relationships between background
and the final nature of companies, a few generalizations are possible
from our data. In the first group, several companies carried over
into prefabrication the organizational characteristics of large con
tracting operations. Some of these tended, after a short period of
true prefabrication, to return once more to the more conventional
patterns from which they had attempted to depart, and although
there might be many other reasons for this return, familiarity with
the old procedures and old friendships undoubtedly exerted their
influence. In general, however, companies developed by engineers
and builders were not wealthy or large enough to carry out a radical
approach to design, even had they wished to do so.

In the second group, particularly among the lumber dealers, the
tendency was to regard prefabrication as a mere refining operation
for the materials handled. Indeed, during the materials shortages
following the war, some never prefabricated houses, but only took

advantage of their favorable supply situation and the regulations of
the Office of Price Administration to charge substantially higher
prices for performing a few additional operations on the materials
as "prefabricators." Others, however, particularly in the major lum
ber supply area of the Pacific Northwest, made use of their experi
ence with distribution and manufacturing methods to bring a genu
ine efficiency to the manufacture of houses.

As for the third group, architects have contributed theories more
often than they have started companies, and, when they have set up
companies, they have often met with difficulties. A few, however,
have been aware of the complexities of operating in the house manu
facturing field and have been able, usually by marshaling other tal
ents about them, to build good organizations.

The fourth group, with experience in manufacturing enterprises,
often had the tremendous initial advantages of well-rounded staffs
and good capitalization; some of them, however, have been impeded
by their attachment to certain materials or by the deficiencies of
their media of distribution. In general they have been characterized
by a willingness to try new materials and designs which might be
well suited to mass production, and for that reason they have been
very important to the industry.

Regarding the remaining three groups, the salesmen, bankers, and
lawyers, the only valid generalization that can be made is that both
their strength and their weakness lay in their emphasis on detailed
organization and salesmanship.

On the whole, the men in top management positions had not been
trained in the industry itself, although a few companies had been
started or staffed by "graduates" of other companies. This is under
standable when it is realized that a man with 15 years' experience
could rightly consider himself a charter member of the industry. A
breakdown of the industry by length of time each company has been
in business will highlight this point, the more so because many of the
older firms were really precutters rather than prefabricators. As of
1947, the age distribution data from 118 of the companies in our
analysis was:

Number of Years Number of

in Business Companies

2 or less 67

3-7 24

8-17 19

18 or more 8

146

An important recent source of trained men has been the various
federal agencies dealing with housing, and there is a trend for in
creasingly large numbers of men to enter the field from the profes
sions and from special courses in the colleges, in the hope of growing
up with the industry.

The size of the staff may be used as a reasonable, if rough, indica
tion of its ability to handle the complete pattern of operations. At
the time of our analysis, at least 50 companies were known to have
staffs exceeding 15 in number, a number probably adequate for the
job. Of those having less, the majority tended to cut out certain
services which they regarded as unimportant; architectural services
were among those most frequently so regarded.

Fortunately, the minimum requirements of the FHA and of build
ing codes have helped to prevent some of the worst errors which might
have resulted from this combination of ignorance and the desire to
keep down costs. A few companies have hired consultants to advise
them on various aspects of their operations, with the result that a
group— very small as yet— of specialists has grown up to serve in this
way. Other producers have allied with independent organizations
which would distribute their output; these distributors were often
land developers as well. Still others have purchased the design, pro
duction, and even procurement and advertising services of a parent
licensing organization; while a few, offering only design or production
ideas, have sought out other organizations with the capital and ability
to take over the rest of the operations.

With regard to the function of research, no prefabricator was
doing what might be called pure research and very few were doing
applied research, although nearly all the 80 largest companies had
staff personnel engaged at least part time in short-range product
development work. There were 51 companies which had part-time
research personnel; 25 had full-time research personnel; and at least
15 had a separate research and development division. Naturally
enough, the companies in the process of getting started were the
more likely to be engaged in concentrated development work, while
those under way tended to abandon research for the more pressing
problems of production and distribution, hoping to return to it when
their volume could support the expense and when they had had a
chance to put their initial designs to a practical test.

147

II. Labor Relations

In discussing labor relations in the industry a distinction should
be made between conditions in the factory and those at the site, for
there are substantial differences. Of the industry as a whole, how
ever, it can be said that labor relations have been generally good.
Our survey found this to be particularly true in the plants, and a
similar generalization has been made by PHMI, which reported that
"relations between employee and employer have been uniformly
good."1 True, there have been instances of restrictive practices,
but the testimony of the manufacturers seems to indicate that re
ports of labor opposition to prefabrication have been magnified out
of all proportion, and certainly since the end of the recent war there
seems to be little justification for the accusation that organized labor
as such is holding up the development of the industry. In part,
labor's attitude stems from the pledge to cooperate made by unions
during the Wyatt program; once good relationships were entered
into, most unions found it to their own advantage to continue in
this way. Of some import have been such factors as the recent high
level of construction activity and the plentiful supply of construction
jobs. Probably more important have been the facts that most of the
producing units are relatively new and small, and that the volume
of the industry as a whole has not yet been such as to attract special
labor interest.

Labor Relations in the Plant

A. Unions

The extent to which the industry had been organized at the time
of the survey was difficult to determine because the situation was in

1 Quoted from testimony by Harry Steidle before the Joint Committee on
Housing of the 80th Congress, January 14, 1948. Mr. Steidle was referring to
conditions in the plant. Austin Drewry, then President of PHMI, described
employer-employee relations as "excellent" in his opening address, Fifth Annual
Meeting, PHMI, Chicago, March 1948.

148

a state of flux. The indications were, however, that in 1947 at least
two -thirds and probably three-quarters of the industry was union
ized, measured either by number of companies or by number of
employees. The AFL had organized about seven times as many
plants as the CIO, there being several unaffiliated unions also.
Among the AFL shops, the most prevalent union was the United
Brotherhood of Carpenters and Joiners of America. Several plants
were organized by such affiliates of this brotherhood as the millmen,
boxmakers, or lumber and sawmill workers, who had less of a craft
background and lower wage rates than the carpenters.

With the advent of increasing industrialization in house manufac
ture, and particularly with the increasing use of materials not tra
ditionally handled by union members in the housebuilding trades,
the CIO began to organize prefabrication plants. For a while it
made some progress, aided by manufacturers who sought an end
to the restrictive practices of craft unions, and occasionally by the
circumstance that an existing union organization might be carried
over from another enterprise which had previously occupied the same
plant.2 More recently, however, the CIO has lost ground, at least
relative to the AFL. A substantial obstacle in its path has been
the trouble sometimes encountered in the field where AFL labor used
for erecting the house refused to handle material made by CIO
labor. Another obstacle was the task of organizing trades such as
those of the plumbers and electricians which have traditionally been
organized along craft rather than industrial lines and have been
saturated with craft attitudes. A not inconsiderable factor in explain
ing the relative halt in the CIO's organizing drive has been the
failure of some of the prefabricators using metal, many of whose
plants the CIO had organized.

One important effect of the CIO's organizing drive, however, was
to provoke the AFL into meeting the challenge. AFL unions have
entered into a number of agreements which indicate the AFL's deter
mination to retain its position in the residential construction field,
even in its most industrialized aspects. Contracts with some of the
larger prefabricators such as Gunnison, National Homes, Pease, and
Lustron, and with Borg-Warner are examples. The last two cases
illustrate the special effort made by many prefabricators to secure
union support. Borg-Warner went to the plumbers' union at an
early stage and secured the endorsement of the international office
on the idea. This company further went to the point of employing a

2 For instance, two prefabricators were organized by the Industrial Union of
Marine and Shipbuilding Workers of America, CIO, for substantially this reason.

149

man with a craft union background to handle relationships with the
master plumbers, through whom it distributed the Ingersoll Utility
Unit, and to participate generally in the labor relations between
plant and field. In the factory, contracts were made with the
plumbers, sheet-metal workers, and electricians. The Lustron man
agement also requested unionization from the start; an agreement
was made with three AFL unions to cover the whole building process,
and it may stand as an example of the growing trend towards reduc
ing the number of craft unions engaged in one building job; in this
case lathers, plasterers, and painters were eliminated. The contract
was made with the international offices of the carpenters, plumbers,
and electricians in November 1947 and was featured at the 1948
AFL convention as a sign of labor's willingness to cooperate. It
provides for a union shop, for uninterrupted production and effi
cient erection at United States and Canadian sites, and for the
avoidance of jurisdictional disputes by limiting the number of crafts
and by including the pledge of the international office to advise
locals and enforce the contract.

B. Wages

As might be expected, wage rates in the plant generally appeared
to be lower than those in the field for several reasons. First, the
order of skills required is lower. Second, there is longer and steadier
employment and therefore the likelihood of better annual take-home
pay, both because of less seasonality in the volume of work and
because of a lower rate of turnover among jobs. Third, working
conditions are better. A rough measure of the difference which
existed between factory and field wage rates is given by the fol
lowing figures: average earnings per hour of employment in 38
prefabrication plants working in wood in July 1947 were $1.14,3 while
average earnings for carpenters in all private building projects at
the same time were $1.58.* In industries somewhat allied to pre
fabrication, however, factory wages were lower: furniture and fin-

3PHMI Survey of Prefabrication Activity, 1947. (Actually this figure is a
bit high since it includes a small amount of overtime earnings.)

4 Monthly Labor Review, 65 (October 1947), 509. The comparison cannot
be exact since the averages conceal rather large geographical variations which
are not weighted equitably for purposes of comparison. Prefabricators usually
more closely approach project rates of their own areas.

ished lumber products, $1.059 per hour; lumber and basic timber
products, $1.033 per hour.5

Some elements in the carpenters' union have gone on record
against this differential in wage rates,6 but, notwithstanding this,
there seems to be a trend towards paying union labor in the factory
at a lower hourly rate than members of the same union receive in the
field. (Of course, the annual pay may be the same, or higher.) In
the Kaiser Community Homes plant at Los Angeles, in February
1947, the several hundred plant employees came under a specially
negotiated contract calling for an AFL closed shop. All plant men,
with the exception of about 15 painters, came under the agreement
made with the International Office of the United Brotherhood of
Carpenters and Joiners of America, rather than with any local.7

Inasmuch as a good many plants must still hire hands on a sea
sonal basis, they are not yet entitled to contend that wage rates
should be lower on the grounds that they offer stabilized employ
ment, although other arguments may be valid. Where prefabricators
have demonstrated the stable nature of their operations, the car-

5 Ibid., p. 500.

6 ". . . Therefore, be it resolved that the United Brotherhood of Carpenters
and Joiners of America immediately put into force and effect the prevailing con
struction carpenter's wage scale for all work performed within the pre-fab, pre-cut
and mill industry which is normally performed on the job site by construction
carpenters."

Resolution No. 13, approved and endorsed by the San Francisco Bay Dis
trict Council of Carpenters, the California State Mill Committee, and six Cali
fornia locals, Proceedings of the Twenty-Fifth General Convention of the United
Brotherhood of Carpenters and Joiners of America (Lakeland, Florida, April 22-
30, 1946), p. 396.

7 Certain jobs in the plant required journeymen carpenters' rates: those which
would ordinarily be carpenters' work if done in the field. Jobs such as nailing
framing members or cutting rafters earned a basic field rate in the plant, for
instance, but the slightly different circumstances yielded a somewhat lower final
rate of pay than the outside rate. Men in the plant got vacations with pay,
however, while field carpenters did not. On the other hand, field carpenters
received double pay for overtime up to the first four hours (45-hour week, gen
erally). Millmen, affiliates of the carpenters, but included under the one agree
ment like all other men in the plant, ran the jointers, bandsaws, etc. Those who
stapled plywood onto framing, ran the portable sanding machines, and did other
semiskilled tasks of a repetitive nature received 33# per hour less than journey
men's rates. Others who came under the carpenters' agreement were the cabinet
makers, the lumber handlers and millhelpers (members of the Lumber and
Sawmill Workers local affiliated with the Brotherhood), clerks, and checkers.
Plant foremen received 12^ per hour more than the journeymen carpenters, but
5tf per hour less than a comparable job in the field. In general, it appears that
wage rates paid in the plant were slightly lower than those paid at the site.

penters, through their international office, have in some cases entered
into contracts which establish the principle of differentiation be
tween field journeymen and factory journeymen rates of pay for
performance of the same type of work. Two trends thus seem evi
dent: a decrease in the number of unions with which the prefabricate!*
has to deal, and a growing acceptance by trade unions of different
wage rates in different conditions of employment, even for the same
type of work.

Nor are these the only signs of change introduced by the pre-
fabricator into the whole pattern of industrial relations in the house
building industry. There is the growing acceptance of many of the
welfare provisions which have for a long time been incorporated
into union-management contracts in other industries, such as paid
vacations, health insurance, and retirement plans. PHMI found in
1947 that 33 member companies had one or more of the following: 8

Number of Number of

Companies Employees

Life insurance 12 1,146

Health insurance 20 2,357

Paid vacation 26 2,624

Retirement plans 3 497

In addition, PHMI found that 15 companies employing 1,347 work
ers had wage incentive or bonus plans of one sort or another— again
indicating that old patterns in the building trades were being changed.9
In the past few years there has been considerable discussion con
cerning the guaranteed annual wage. The CIO has taken a strong
position in favor of such a plan for the building industry,10 while the
AFL has voiced equally strong opposition, holding that building is
clearly a field in which a guaranteed wage plan cannot be made to
work.11 The recent experience of prefabricators with the problems
of stabilizing sales, procurement, and production has quite naturally
led them to consider any such scheme a grave risk. While the issues
involved are complex, it does seem evident that the guaranteed an-

8 PHMI Survey of Prefabrication Activity, 1947.

9 Loc. cit.

10 Testimony of R. J. Thomas, President, United Automobile, Aircraft and
Agricultural Implement Workers of America ( CIO ) ; Chairman of the CIO Hous
ing Committee, given before the Senate Special Committee on Post-War Eco
nomic Policy and Planning, 79th Congress, 1st Session (Post-War Economic
Policy and Planning, Part 10, pp. 1678-9).

11 William Green, "Your Postwar Income," American Federationist, 52 (April
1945), 32-3.

152

nual wage is still far from realization in the industry, because of the
AFL's dominance in prefabrication, the general opposition to the
scheme by the building-trades unions, the prefabricated struggle to
overcome seasonality (which has thus far been but partially success
ful), and his often insecure financial position.

C. Restrictive Practices

Despite the generally good labor relations in the plants, there
have been instances of restrictive practices on the part of unions
at this level. The testimony of certain manufacturers who related
particular incidents must be weighted more heavily than the flat
denial of such practices by some union spokesmen. Six companies
stated that they were prevented from spraying paints in their plant
because of union opposition, while others had spraying time severely
limited. Whether the union's usual objection that spray guns are
not safe from a health standpoint was adequately met in these cases
is not known, but paint spraying can be made safe in a factory, and
it is a fact that spray guns were known to be in use in seven AFL
plants. Some 15 companies stated positively that they were pre
vented from prefabricating plumbing of any sort because of union
opposition. Many more were probably affected. In the case of
plumbing, however, it is difficult to separate union opposition from
what might better be termed resistance to a change in conventional
plumbing material distribution methods, since the two are usually
tied together, and from the effects of local building-code regula
tions. Twenty-seven companies stated that they had refrained from
prefabricating plumbing because of a combination of these factors.
Master plumbers have a natural interest in opposing the prefabrica
tion of plumbing, since its logical course is to reduce or eliminate
their sales of fixtures and supplies in connection with their installa
tion work. Prefabricators were able in some cases to make arrange
ments with master plumbers to fabricate plumbing assemblies in the
plant through what amounted to a royalty agreement with the
master plumbers. On the other hand, at least 27 companies were
known to be preassembling plumbing and six companies were known
to be precutting it, which indicates that a new pattern is evolving
and that through persuasion and compromise some of the opposi
tion is disappearing as time goes on.

153

There have also been cases where, because of definite opposition
from the electricians' union, wiring was not preins tailed. However,
it is often difficult to know just how the opposition is exerted, and
just what connection, if any, the unions have with code provisions
which protect their special interest. A not uncommon practice, for
instance, is to resist prefabrication indirectly through a general code
requirement such as that all wiring installations be field inspected.
If such a provision is literally interpreted, it can prevent the pre-
installation of wiring in panels which have both interior and exterior
surfaces applied in the factory. Furthermore, some codes require
rigid metal conduiting for electrical installations, while others require
flexible cable, thus making it impossible for the prefabricator to stand
ardize his installations.

In some areas, according to the prefabricators, union officials have
made purposefully unreasonable demands as to the number of skilled
workers and the general wage rates to be agreed to by management
before production could start, in order to prevent companies from
establishing themselves in the house manufacturing field.

No prefabricator reported having trouble in his plant from juris-
dictional disputes. There may have been some loss in the potential
efficiency of the labor force on occasions when the union contract
would not permit a prefabricator to shift workers from one task to
another, but no particular instance of this sort was mentioned. Gen
eral Homes, Inc., which was organized by the CIO, did emphasize
the importance of the provision in its contract which permitted any
man in the plant to shift to different tasks at different times as the
situation might demand. Such a provision might well help to stabi
lize the factory labor force, and might be particularly important in
the field, where flexibility would help to avoid delays and would per
mit the more efficient use of manpower. This sort of thing has been
resisted by the craft unions, but the CIO has strongly supported it.
The following letters are an indication of this attitude:

Since we are an industrial union, we have no difficulty with jurisdictional
matters. We apply the same policy to those of our workers engaged in
prefabrication of homes as we do to those engaged in the shipbuilding in
dustry.

. . . We do not oppose any device to expand the average productivity
of the individual work. However, we do insist the economic result of
increased productivity be shared by the worker, as well as by management
and the consumer. . . ,12

12 John Green, President, Industrial Union of Marine and Shipbuilding Work
ers of America, in a letter to the Bemis Foundation, June 10, 1947.

Members of this union are engaged in various types of lumber pre-
fabrication.

We are not opposed to labor-saving devices, provided the workers are
given a fair share of the added production which can be achieved through
the use of such devices. . . .

We are an industrial union in the logging and woodworking industry,
and have no internal jurisdictional problems like the building trades craft
unions. . . ,18

Generally speaking, union efforts to restrict plant prefabrication
have been minor in extent, and there is evidence that they are be
coming steadily even less important.

Labor Relations in the Field

Labor relations of prefabricators in the field are much the same
as in the rest of the housebuilding industry. They are, of course,
tied up with the prefabricated marketing system, for when he
undertakes to do his own erection, he is likely to use a different
form of labor organization from that used by a dealer-erector. In
fact, dealer-erectors have handled most of the erection work, and their
labor relations have been typical of the small- or medium-sized
builder and have involved a number of the same AFL unions at
their regular hourly wage rates. On the other hand, General Homes
planned to carry out its own erections with CIO labor under the
contract mentioned above, an arrangement sought partly out of the
fear that AFL labor would refuse to handle the job.

Such a fear was not wholly ill-founded. There have been a num
ber of instances in which AFL unions have opposed the erection of
houses fabricated by another AFL organization or by non-union or
CIO workers. These date back at least as far as the well-publicized
occasion in 1940 when a gang of AFL men attacked a CIO erection
crew working on a Gunnison house in East St. Louis, Mo. In the
past few years there have been other incidents.

In early 1947, The Green Lumber Company, which had established
a CIO shop through a recent election, sold some 200 houses to a
builder in Jackson, Miss. When the builder tried to hire AFL men
to erect his houses, the business agent refused to allow his men to
handle the job. Although both the National Housing Agency and
CIO officials appealed to him, the sale had to be canceled.

18 J. E. Fadling, President, International Woodworkers of America, in a letter
to the Bemis Foundation, June 9, 1947.

755

The Scott Lumber Company, producing Scott Homes, presented a
somewhat different case at about the same time. The plant had been
organized as a closed shop by the AFL carpenters, but the dealer
in Dunkirk, N. Y., ran into strong opposition from local 689 of the
carpenters. The local's position was that its members wanted to
"build houses, not erect them," that there was plenty of labor and
materials locally available to do the job, and that the Scott men,
albeit of the same brotherhood, were performing in the factory at
Wheeling, W. Va., the work they felt themselves entitled to do.14

The Harnischfeger Corporation also ran into trouble in 1947 when
it refused to agree to a preferential hiring clause in a contract with
the carpenters. The union ordered its members to discontinue erec
tion of the houses and sent letters to its membership advising that
neither they nor members of any other building-trades union erect
houses which did not bear the label of the United Brotherhood.15
The carpenters even went so far as to direct their locals and district
councils to adopt a by-law that

No member will use, handle, install or erect any material produced or
manufactured from wood not made by members of the United Brother
hood.16

It should be noted that the above practices might well be found
to be illegal under the secondary boycott provisions of the Taft-
Hartley law.17

There have been other forms of union obstruction in the field,
such as refusal to handle certain prefabricated elements— plumbing,
preglazed sash, prehung doors. On occasions there have been delays
caused by jurisdictional disputes, most frequently in the case of the
erection of a metal house where no clear precedent had been estab
lished. The William H. Harman Corporation encountered one such
instance; 18 and on occasion there has been a more general and very
understandable opposition, such as that voiced by William J. Mc-
Sorley, General President of the Wood, Wire, and Metal Lathers
International Union, AFL:

14 Dunkirk Evening Observer, March 29, 1947, p. 1.

15 Labor Relations Reporter, Vol. 20, no. 51 (October 27, 1947), 395-6.

16 Minutes of the Meeting of the General Executive Board, Lakeland, Fla.,
January 16, 1947, The Carpenter, LXVII (March 1947), 21.

17 Labor-Management Relations Act, 1947 (Public Law 101), effective June
23, 1947, Section 8 (b) (4) (A).

18 The New York Times, August 20, 1948, p. 18. Jurisdictional disputes are
also illegal under the Taft-Hartley law, but delays can occur without there be
ing a strike and without the case coming to court.

156

... I desire to say . . . that most of the prefabricated houses are de
signed and built without any lathing and plastering in them. . . . This of
course is one of the principal reasons why we are opposed to prefabricated
housing, and some of the reasons are contained in the enclosed pamphlet
which has been issued by the National Foundation for Lathing and Plaster
ing. . . . We believe that all houses that are erected for the purpose of
housing human beings should be lathed and plastered in a proper manner,
so as to protect sanitation and health of the inhabitants ... to be candid,
we are not doing anything to promote any program that will have a
tendency to put us out of business. . . ,19

Union opposition in the field has thus been of considerably more
concern than that in the shop. But it should be remembered that
the few cases of union opposition get more publicity than the many
cases of union cooperation. Building is not the only field in which
technological change has been resisted, and experience shows that
adjustments are made in the course of time. Prefabrication, in one
form or another, is a growing reality; the need for housing calls for
production in tremendous quantities; and public pressure will call
for an end to restrictive practices. In view of these factors it does not
seem unreasonably optimistic to summarize that the problem of union
opposition is relatively small and appears to be growing smaller.

III. Financing

A. Capitalization

It has frequently been said in the industry that a successful pre-
fabricator requires about $1,000,000 in capitalization. How many
of them have reached this figure? Very little information on capital
investment in the industry is publicly available. Some manufacturers
decline to reveal such figures, and in other cases the capital invest
ment for the production of prefabricated houses is hidden in a figure
giving the total capitalization of a firm in which prefabrication is but
a subsidiary activity.

Table 1 gives the distribution of capital ratings published in 1947

19 In a letter to the Bemis Foundation, June 2, 1947.

757

Table 1

Capital Rating of Prefabricators

$1,000,000

$500,000

$300,000

$100,OOD

$10,000

$25,000

$50,000

126 Firms from Thomas' Register of American Manufacturers, Ed. 38, vol. 1
(December 1947), columns 7863-5.

158

by one register. It will be noted that information was available for
only slightly over half of the companies listed in the register. Those
which refused information included several of the largest companies.
It should be pointed out, however, that the 13 companies capitalized
at over $1,000,000 represented for the most part investments not solely
or even primarily in prefabrication. Only a few of the 13 started out
as prefabricators and reached the million-dollar class.

PHMI estimated that at the end of 1947 the 80 or so companies that
were actively engaged in prefabrication represented a total capital
investment of about $60,000,000, with an additional $36,000,000 in
vested in the industry's dealers.20 These figures, too, would indicate
that there may be an appreciable number of firms in the industry
capitalized in excess of $1,000,000.

B. Sources of Investment Capital

Only three prefabricators are known to have raised their capital
through public stock subscription: Anchorage Homes, Inc., General
Panel Corporation of New York, and William H. Harman Corpora
tion. In the majority of cases, capital has been obtained privately,
usually through individuals, sometimes through parent organizations,
but seldom through financial institutions. It is more or less to be
expected in an industry such as this where risks have been high that
the banks would be of only minor assistance. Only 14 companies
indicated that they had gone to banks for long-term capital loans,
while 10 companies reported that they had experienced difficulties
with banks; most companies reported no dealings at all with banks
in connection with long-term capital requirements.

A large segment of the industry has been financed by parent cor
porations of one type or another. Among the prefabricators active
at the time of our survey, many owed either their original formation
or much of their capitalization to large industrial enterprises. Partly
because of this parenthood, these were some of the best-known names:
Gunnison Homes (United States Steel Corporation); Stran-Steel Arch
Rib Homes (Great Lakes Steel Corporation); Wingfoot Homes
(Goodyear Tire & Rubber Co.); Butler Homes (Butler Manufactur
ing Company); Kaiser Community Homes (Henry J. Kaiser); Ther-
mo-namel Houses (Higgins Industries); Lustron (Chicago Vitreous

20 Austin Drewry, President, PHMI, Opening Address at Winter Meeting, De
cember 15, 1947.

159

Enamel Product Co.); and P & H Homes ( Harnischf eger Corpora
tion). Such companies also had some of the largest and best-
equipped plants in the industry.

Other prefabricators have been financed in large part by large con
tractors and builders, as, for example, Johnson Quality Homes, Inc.,
by John A. Johnson Contracting Corporation, and Kaiser Community
Homes by Fritz B. Burns. Still other companies have been financed
by parent lumber or plywood organizations, such as Prenco by
C. D. Johnson Lumber Corporation, Hayward Homes by Hayward
Lumber and Investment Co., and General Timber Service by Weyer
haeuser Timber Co.21

Some companies have financed part of their operations through a
licensing system under which they receive royalties from licensee
manufacturers operating in various localities. Ivon R. Ford, Inc., had
some nine licensees at the time of the survey, and American Houses,
Inc., had six, in addition to its own three plants.22

There have been numerous enterprises which failed to get into
production for lack of venture capital. While this might be true of
any enterprise, it is probably harder to attract venture capital to
new methods of housebuilding than into most other fields, and per
haps rightly so. Part of the explanation lies in the mass of obstacles
which the innovator in this field faces in the way of restrictive prac
tices, codes, consumer resistance to change, and so forth— a list
which has been enumerated many times. Another part lies in the
extent to which aspects of building permeate a vast range of institu
tions: family, neighborhood, city government, public utilities, or
ganized labor, big business, real estate, financial institutions. What
ever the causes, and there are more than a few, housebuilding has
been dubbed "the industry capitalism forgot" 23 and has been singled
out frequently as that industry most in need of the sort of revolu
tion that has characterized the history of capitalism. Raising venture
money has not been made easier by a number of well-publicized
failures in prefabrication, especially recently, even though an anal
ysis of the proposed patterns of operations would have revealed from

21 Very indirectly, several other companies were related to large capital.
Many Baldwin Locomotive officers were interested in Harman; Consolidated
Vultee decided not to back a house, but some of its officers were associated with
Southern California Homes; and the Ibec house is a venture of Nelson Rockefeller.
The ultimate decision of Beech Aircraft not to produce for Fuller was a major
blow to Fuller Houses.

22 The license arrangement was perhaps most extensively used just before the
war by Precision-Built Homes Corporation which at that time had 67 licensees.

™ Fortune, XXXVI (August 1947), 61.

160

the start that failure was very likely in most cases. Consequently
the companies having the most radical ideas and generally involving
the greatest risks have had even more difficulty with financing than
might be expected. Perhaps the most spectacular case of this sort
was the Fuller house, an enterprise attended by much notoriety be
cause of its boldness and novelty, but one which never got under
way because of failure to attract enough risk capital.24 Other some
what more conservative companies have also had to struggle to get
private financing— Lustron, General Panel Corporation of California,
Reliance Homes, and Southern California Homes are examples, each
one of which proposed major innovations. •

One result of this situation has been a debate, inside the industry
and out, and often quite heated, as to whether a very large initial
investment is a necessary condition for success in prefabrication and
if so whether the government should take an active part by securing,
or even making, such an investment. On one side have been the
older and more conventional members of the industry, usually work
ing in wood, who have held that the industry would grow of itself
if only given the chance, that no huge investments were needed, par
ticularly if they had to be government sponsored, and that no special
favors were required, but only a minimum of government regulation
of sufficient stability to make planning by business possible. On
the other side have been many of the newer and more unconven
tional companies which have argued that thus far prefabrication has
not made good on its promise of cost reduction, that revolution, not
evolution, is necessary, that houses can be mass produced at really
low costs only by an enterprise which represents a complete dis
continuity with the past in both the nature and the scale of its opera
tions, and that in a housing emergency the government should take
an active part in encouraging such ventures. By and large these
divergent opinions were represented respectively by the Prefabri
cated Home Manufacturers' Institute and the National Association
of Housing Manufacturers, but were by no means confined to them.
The latter philosophy lay behind the Wyatt program and in the
somewhat less active role the government has played since the
Veterans' Emergency Housing Program ended. In any case, the gov
ernment has become an important factor in the financing of the
industry in recent years.

The sale or lease of surplus war plants to prefabricators is one
direct means by which the government assisted certain firms in estab-

24 See "What became of the Fuller house," Fortune, XXXVII (May 1948), 168.

161

lishing themselves. The Housing Expediter was empowered by the
Veterans' Emergency Housing Act25 to direct that certain surplus
production facilities be disposed of for use in the manufacture of
housing. Nine prefabricators are known to have acquired plant
facilities in this way, several of the factories being very excellent
buildings once used for aircraft production.26

More important to the capitalization of the industry have been the
three financial mechanisms involved in the government program:
loans, market guarantees, and the insurance of loans made by private
institutions.27 The last two are concerned more with working capital
than investment capital and are discussed later in the section on
credit. The loan program developed out of a background which had
seen the wide use of government powers in times of national defense
and war, and out of legislation that extended some of these powers
into a time of drastic housing emergency. Under the provisions of
the Veterans' Emergency Housing Act the Housing Expediter was
given the authority to direct the RFC to make loans to prefabricators.
Early in the history of this program there were a number of disputes
between the RFC, which declined to make loans that it considered
unsound, and the Office of the Housing Expediter, which held that
the risks were not as great as imagined and that in any event the
housing emergency justified such risks. The nature and outcome of
these disputes were partially responsible for Wyatt's resignation as
Housing Expediter; 28 however, the RFC had made 20 OHE-sponsored
loans to prefabricators by June 1, 1948. These loans totaled $38,-
290,000, and, as of that date, disbursements had been made to 12
of the companies in the total amount of $9,565,000.29

25 Public Law 388, 79th Congress, approved May 22, 1946.

26 For instance, Lustron obtained part of the Curtiss- Wright plant in Co
lumbus, O., and General Panel part of the Lockheed plant in Burbank, Calif.
Source: War Assets Administration, Office of Real Property Disposal, June 1947.

27 It should also be remembered that another type of government assistance
to prefabricators was the priorities and allocations program through which, at a
time of critical postwar shortages, materials were channeled to them.

28 December 4, 1946.

29 Source: RFC records to June 1, 1948, reviewed by the Bemis Foundation.
Of the 8 companies to which no disbursements had been made:

1 loan was outstanding.

7 loans had been canceled. Of these:

4 companies abandoned plans.

1 company obtained financing from other sources.

1 company failed to raise necessary equity.

1 company withdrew application.

162

The largest single loan was the initial loan of $15,500,000 to the
Lustron Corporation, made only after a considerable period during
which the matter was extensively debated. At the time that the
formation of Lustron was announced $840,000 in private capital had
been raised.30 The RFC subsequently made loans to Lustron which
eventually more than doubled the initial amount, by its own decision
and not under direction from the OHE, which was later functioning
in a liquidating capacity only. Thus the most heavily capitalized
enterprise in the industry, one several times bigger than the next
largest firm, was almost entirely financed by the government.

The principle of government loans to prefabricators was extended
by the Housing Act of 1948, passed by the Special Session of the 80th
Congress.31 The Act authorized the RFC to make loans for the
production of prefabricated houses or components or for large-scale
site construction, but if such loans were used for the purchase of
equipment, plant, or machinery the loan was not to exceed 75% of its
purchase price. Such loans were not to exceed $50,000,000 out
standing at any one time, and were not to be made if financing was
otherwise available on reasonable terms.

C. Credit

According to some, the most important and least understood prob
lem facing prefabricators is that of credit. While this may be an
extreme point of view, it is nonetheless true that obtaining credit has
been a crucial question for many firms, particularly in the steady and
continuous flow that may be required throughout every phase of the
housebuilding process— by the prefabricator to pay for raw mate
rials, labor, and other costs of production; by the dealer to pay the
prefabricator for the factory package; and by the homebuyer to pay
the dealer for the completed house. In the production process the
sums involved tend to be very large, and many prefabricators cannot
finance their operations without resorting to working capital loans
of one sort or another.

The total investment of a prefabricator who undertakes to pro
duce 100 house packages at $4,000 per package is $400,000. If
these houses cannot be sold to a dealer for cash, the prefabricators

30 The New York Times, November 1, 1947, p. 22.
81 Public Law 901, approved August 10, 1948.

163

capital will be tied up in them, and he will soon have to cease pro
duction. Similarly, the dealer cannot use his capital again until the
customer pays him for the finished house. It is not enough that
financing be available at all stages; it must be available without
delay, so that the flow of funds will proceed at a pace with the
flow of materials and fabricated products. The last two stages, re
lating to dealer credit and consumer credit, are discussed in the
chapter on marketing, leaving only the credit which is extended to
the prefabricator for working capital loans to be treated here, al
though all three are interrelated.

First of all, it should be pointed out that not all prefabricators
have had a problem in obtaining working capital. Many companies
have had no need to borrow for this purpose, either because they
have steadily accumulated sufficient capital for their scale of opera
tions, or because they have large parent concerns which make such
capital available to them. Other firms were able to obtain credit
from their materials suppliers, especially in cases where the pre
fabricator had previously established contacts with them in some
other type of building enterprise. Most of the older members of the
industry had lines of credit with the banks. Thus it was primarily
the youngest firms, and particularly those which planned to commence
operations on a large scale, that encountered difficulty. Not infre
quently these were regarded as risky ventures, and the problem was
therefore to earn the confidence of the banks. The bankers expected
these firms to prove themselves through successful operations over a
period of time, but how were they to get started?

One device which was designed in part to meet this problem was
the guaranteed market contract, under which it was hoped to re
duce the risk attending a new prefabrication venture by having the
government act in an underwriting capacity. The Veterans' Emer
gency Housing Act authorized the RFC to guarantee markets for
prefabricated houses to the extent found necessary by the Housing
Expediter in order to assure a sufficient supply for the Veterans'
Emergency Housing Program, but the number of houses covered by
the outstanding guarantees was at no time to exceed 200,000, nor
was the net loss to the government to exceed 5% of the total guarantee
undertaken. A number of criteria were set forth: guarantees would
be of temporary duration, would be pointed towards low-cost prod
ucts, would not cut into the market for conventional houses, and
would be awarded only after rigid tests on the house and a demon
stration of ability to perform by the prospective producer. In brief,
the contracts specified a production schedule and provided that if

164

the prefabricator was unable to sell what he had manufactured,
the units would be purchased, subject to certain conditions, by the
RFC. The manufacturer was obligated to repurchase the units from
the RFC before selling any more houses of the same or equivalent
type. Thus it should be noted that the guaranteed market contract
did not provide a market into which could be continuously poured
the output of a prefabricator; it did not offer an opportunity for
operational improvements by absorbing the output during a period
in which changes in design, production, or distribution technique
might be made. Once a prefabricator tendered houses to the gov
ernment, he was essentially forced to halt production. The con
tracts did serve as collateral, however, by certifying that the gov
ernment was ready to buy what could not be sold elsewhere, and
thus they enabled some companies to obtain loans for much needed
capital.

Of 74 companies which applied for market guarantees, 20 received
contracts, all terminating December 31, 1947.32 The contracts guar
anteed the market for 61,696 units out of a total original scheduled
production of 90,596, and involved a total liability of $195,833,708.33
Actually, however, fewer than 3,000 houses 34 were produced under
these contracts, a disappointingly small total which reflects the fact
that many of the companies did not get into production before their
contracts were terminated or, in some cases, canceled by mutual
consent.35 The net loss to the government was about $3,000,000,36
about 1%% of the total liability and well below the specified limit,
but in light of the production that resulted, the program can hardly
be called anything but a failure. A redeeming point was its assist-

32 The balance of 54 did not receive contracts for various reasons, including
the following:

( 1 ) Not enough experience.

(2) Insufficient equity to qualify for RFC loan— therefore had no funds.

(3) House not technically acceptable.

(4) Design too costly, used too much critical material.

(5) Unable to obtain plant or equipment.

(6) Showed only initial interest— did not follow up with necessary papers.

83 Source: OHE official records, reviewed June 1, 1948, by the Bemis Founda
tion.

34 Source: loc. cit.

35 The magazine Business Week (December 11, 1948), p. 25, covering the
marketing and finance problems of prefabricators, stated that only six out of 32
companies which secured guaranteed market contracts or loan agreements through
RFC were still turning out houses.

36 Source: OHE estimate, given to the Bemis Foundation, June 1, 1948.

ance in the formation of several enterprises which may have a
stimulating influence on the development of the industry.

The underwriting of loans was the third of the financial mechanisms
by which the government sought to help prefabricators obtain capi
tal. This program was initiated on July 1, 1947, when Congress, by
amendment of the National Housing Act ( Section 609 ) , provided for
federal insurance of working capital loans. These could be for as
much as 90% of the necessary current cost of manufacturing the
house (package), exclusive of profit. In principle, this extension of
FHA operations had its counterpart in the FHA Title VI program
for conventional construction under which were insured the construc
tion loans used in financing homebuilding at the site. Since con
struction loans for conventional building were being insured only if
the permanent financing for the home had been arranged, a pro
duction loan under Section 609 was to be insured only if the pre-
fabricator submitted binding purchase contracts as collateral evidence
of sale and ability to pay for houses manufactured with the proceeds
of the loan.

By April 30, 1948, when the original form of the Title VI program
expired, 24 applications for Section 609 loans had been received.
Only one company, however (Housemart, Inc.), had actually ob
tained an insured loan, and this was for the production of 194
houses. Why, it may be asked, were so few houses financed under
this program during the 10 months it was in effect? Part of the
answer to this question lies in the difficulty of judging the technical
merits of an applicant's product by a review of plans and specifica
tions, and by examining and testing a hand-made prototype not pro
duced under conditions to be expected in full production. Part of
the answer lies in the length of time required to investigate all those
other aspects of the applicant's business operations which the FHA
considered it necessary to investigate— the borrower's plant facilities,
financial condition, manufacturing costs, marketing plans, etc. But,
to the largest extent, the answer involves the "binding purchase
contract" which the prefabricator was required to show before he
could obtain a loan. Section 609 did not define such a contract in
specific terms, and certain applicants for loans were led to criticize
the FHA's interpretation of the phrase, which was cautious and con
servative. In effect, the FHA did not wish to be involved in insur
ing the marketability of the houses; it wished to make certain that
they were not being produced for an unknown market. No loan was
approved for insurance unless the dealer-erector involved in the
purchase contract could show that he had the necessary cash in

166

hand or arranged for, which meant that he must have arranged the
permanent financing for the houses before the fabrication process
could start. Many housing manufacturers held that this was an
unrealistic requirement, that it was not practicable to make these
financing arrangements so far in advance of delivery of the houses.

The housing manufacturers feel that FHA will be fully protected if by the
time the houses are to be delivered under a purchase contract, the pur
chaser is required to have the cash for payment on delivery, or to have
financing arranged which assures the payment of the balance due under
the contract.87

It is clear that one issue involved here is the extent and nature
of risk contemplated by Congress when it enacted Section 609. But
whatever the pros and cons of FHA policies in regard to this pro
gram, it remains a fact that, in its original form, it fell far short of
its objectives, and changes were introduced into Section 609 when it
was reenacted along with other elements of the FHA's Title VI
program in the Housing Act of 1948.38

In this new form Section 609 authorized insurance of loans for the
manufacture of prefabricated house packages on the basis of con
tracts (for the purchase of these packages) which provide for pay
ment of the purchase price within 30 days after delivery of the
houses, or payment of 20% of the purchase price on or before delivery
if the institution making the loan to the manufacturer accepts and
discounts a promissory note for the unpaid balance payable within
180 days from the delivery date. In addition to insuring loans to
finance the production of house packages, the new Section included
provisions for short-term financing of dealer-erectors by authorizing
the FHA to insure the lending institution against losses sustained in
accepting and discounting promissory notes of purchasers represent
ing the unpaid purchase price of the packages. These notes could
not exceed 80% of the purchase price, nor could they have a ma
turity in excess of 180 days.

A further feature of great importance was also added: the manu
facturer was permitted to substitute new purchase contracts as security
on the loan in place of contracts which had been performed. This,
in effect, made the principal amount of the loan a revolving fund for

37 ( Our italics. ) Statement by Nathan Wendell, Vice-President of the Na
tional Association of Housing Manufacturers and Vice-President of General Panel
Corporation of California, given before the Joint Committee on Housing, 80th
Congress, 1st Session (Study and Investigation of Housing, Part 5, p. 5,062).

38 Public Law 901, approved August 10, 1948.

167

financing the production of additional houses above the number
stated in the original loan agreement.

IV. Public Relations

Public relations has been an important problem of the prefabricator
for more than a decade. Before the war it was principally a matter
of overcoming consumer prejudice against novelty in the design of
the house. Since the war it has more frequently been a question
of correcting the impression that prefabricated houses are temporary
dwellings which are structurally inadequate.

Public attitudes have exerted their influence not only in consumer
resistance, but also in active and organized opposition to the erection
of prefabricated houses in certain communities. A typical example
was the trouble encountered in 1947 when an attempt was made to
erect in Natick, Mass., a suburb of Boston, a house produced by
Winner Manufacturing Company, Inc., under license from Shelter
Industries, Inc. The house was of modern design and stressed skin
plywood construction. A building permit had been granted and
erection was under way when a group of neighbors, fearing that
their property values would be seriously endangered, brought pres
sure on the building inspector to revoke the permit which he had
already issued. After appeal to a special emergency board which
had been set up in Massachusetts, and a consideration of this appeal
mechanism by the courts, the permit was finally granted. Such were
the difficulties and the character of public opinion, however, that
the company later turned its attention to other areas with a modified
design.

One indication of the importance of public relations to the pre
fabricator is the considerable number of firms, 18, which our survey
found using public relations agencies or counselors. There has been,
of course, a great deal of free publicity given to prefabricators in all
types of communication media, and this is, perhaps more than any
thing else, a reflection of the keen interest of the public in anything
which might help solve the housing problem. Much of this publicity
has been the wildest sort of fantasy, however, and much more has

168

been entirely premature; it has probably done the industry more harm
than good.

Because some unquestionably poor houses have been produced by
prefabricators and because some of the most widely publicized ven
tures have come to naught, many firms have sought to distinguish
themselves from the rest of the field by means of carefully directed
advertising campaigns. Some do cooperative advertising with their
dealers, splitting the cost, so that they can control the content and
quality of the ads. Others seek to avoid identification with prefabri-
cation entirely by disclaiming any resemblance to all that has gone by
that name, by designing and erecting their houses so that they can
not be distinguished from the conventional product, or by employing
such terms as "prebuilt," "pre-engineered," "manufactured homes."
There is, in fact, strong support for the abandonment of the term
"prefabrication" in favor of "house manufacturing" or "home manu
facturing." The use of the seal of the Prefabricated Home Manufac
turers' Institute is another means by which some companies have
sought to create a reputation of soundness for their products. And
a very influential factor in establishing a prefabricator's reputation
for quality is the approval of his house for mortgage insurance by
the FHA.

Mass distribution through brand-name selling is one of the im
portant potential advantages offered by prefabrication. This is recog
nized in varying degrees by most prefabricators and has been heavily
stressed by a few. As the housing market has changed, and as the
scope of the marketing problem has come to be recognized in the
past few years, there has been increasing emphasis on the selling ef
fort required and on the advertising that must be a part of the dis
tribution pattern. Most firms engage in some form of advertising
in addition to their descriptive brochures, usually in local newspapers
and in trade journals; a few firms, like Lustron, Adirondack Log Cabin,
and National Homes, have also done magazine advertising on a nation
wide basis. But it is probably a fair generalization that the develop
ment of brand-name selling through advertising has not yet been
carried beyond the initial stages by the vast majority of prefabricators.

169

V. Trade Associations

There are at present two trade associations functioning in the in
dustry, the Prefabricated Home Manufacturers' Institute39 and the
National Association of Housing Manufacturers.40 These differ quite
markedly in their membership, policies, and activities.

A. Prefabricated Home Manufacturers' Institute

The need for an association of prefabricators in 1942 led a half-
dozen of the industry's pioneers to form the Prefabricated Home
Manufacturers' Association. In 1943 the Association renamed itself
an Institute, there being 12 charter members. By January 1946 mem
bership included 30 companies, and in the boom days of early 1947
it reached a peak of 67. In mid-1948 membership was stabilizing at
46. While membership is not limited to users of specific materials,
PHMI is largely comprised of those firms which work in wood (in
cluding plywood), and which have approached most aspects of pre-
fabrication with what might best be called a conservative attitude.
Among the companies which have been most active in the organiza
tion are Gunnison Homes, Inc., National Homes Corporation, Pease
Woodwork Company, Inc., American Houses, Inc., Southern Mill
& Manufacturing Co., The Green Lumber Company, Houston Ready-
Cut House Co., Crawford Corporation, Ivon R. Ford, Inc., Page and
Hill Co., Harnischfeger Corporation, and Johnson Quality Homes,
Inc. The PHMI staff includes a manager, a public relations man,
and a statistician and cost accountant. Harry H. Steidle, who heads
the staff, was for five years Washington representative for the Douglas
Fir Plywood Association and active in other trade association work
before joining PHMI. Some years previously, he was Assistant Chief

39 908 20th St., N.W., Washington, D. C.

*o 1028 Connecticut Ave., N.W., Washington, D. C.

Although the Douglas Fir Plywood Association is not, properly speaking, in
the industry, it should be mentioned here because of its promotional activities
in behalf of prefabrication since 1938. The Association regards prefabricated
houses as an important long-run market for plywood and has published several
booklets to further this type of construction.

170

of the Division of Trade Standards in the National Bureau of
Standards.

PHMI maintains about a dozen standing committees, the most im
portant of which deal with industry promotion, marketing, employer-
employee relations, technical problems, accounting and statistics, ma
terials, and government relations. The accounting and statistics com
mittee has made some progress towards having a uniform cost-ac
counting system adopted by member companies, while the technical
committee has developed a set of performance standards for prefabri
cated houses which was published as Commercial Standard 125-45 by
the National Bureau of Standards.41 In addition, the technical com
mittee has worked with building-code officials in various sections of
the country in order to reduce the code conflicts facing the industry.

In its public relations role, PHMI furnishes information to the press
and interested individuals and represents the membership at builders'
conventions and similar meetings. The Institute's advertising cam
paign, with its use of a seal and its emphasis on quality standards,
has been an important part of this program. Members receive a
weekly newsletter which presents an excellent summary of housing
activity and pertinent legislation, occasional generalized reports on
the operations of member companies, and other news of interest.
Conventions are held several times a year and provide an opportunity
for the exchange of information.

Naturally, an important function of the Washington office has been
to represent the interests of member companies and, when appropri
ate, the industry as a whole, in the various federal agencies connected
with housing and before committees of Congress. During the Vet
erans' Emergency Housing Program, when the government allocated
materials and controlled prices, this function was particularly im
portant. The attitude of PHMI towards a government program for
prefabricators, however, has rather consistently opposed special aids.
The organization was against many elements of the Wyatt program
on the grounds that they would bring into existence many get-rich-
quick firms which could not last but which would impair the indus
try's reputation and credit standing. This attitude has been strength
ened recently with the failure of inexperienced government-financed
firms, which resulted in the general loss of confidence in the industry
in some banking circles. While PHMI has opposed guaranteed
markets, RFC loans, and the Housing Act of 1949, it has fought for

41 A second edition, Commercial Standard CS 125-47, was published in No
vember 1947.

171

liberal mortgage financing under Title VI and for stability and con
sistency in federal housing legislation so that long-range plans could
be made by those in the business of providing homes.

B. National Association of Housing Manufacturers

Founded in February 1947, NAHM from the start emphasized the
use of modern methods and improved building materials and tech
niques. Its efforts have been largely directed towards the new firms
in the industry which have been using unconventional materials and
new designs. The organization was not designed to serve as a public
relations front, and so it has avoided publicity as much as possible,
although it has testified at hearings.

The primary purpose has been to help secure the necessary legisla
tion and regulations to make available the government assistance
which these companies require, including loans for working capital
and the marketing of houses, priorities and allocations of materials,
and mortgage financing for the completed houses. In this connection,
NAHM representatives have testified before various Congressional
committees as well as committees and agencies within the executive
branch of the government, such as, for example, the Office of Indus
try Cooperation of the Department of Commerce, where the volun
tary allocations program has been administered. NAHM has been
of considerable influence in securing the legislation and assistance re
quired by the industry, but in general it has remained in the back
ground.

The membership has varied from time to time, and a number of
non-member companies have participated in the meetings. Attend
ing these have been as many as 15-20 companies, among which were
a few producers of new-type housing materials. Among the com
panies which have been most active in the Association have been
Lustron, whose President, Carl Strandlund, was the initiating force
behind the Association and has been its President from the beginning;
General Panel, whose President, Abel Wohlstetter, is the Vice-Presi-
dent of the Association; and Reliance, whose President, Harry Nagin,
is also a Vice-President of the Association. Counsel is David L.
Krooth, former General Counsel of the National Housing Agency and
of the Housing Expediter.

NAHM thinks of itself as representing the producers of industrial
ized or machine-made housing, rather than the prefabricators, who,

172

it holds, are for the most part still working in conventional materials
in the conventionally inefficient way. Its policies and actions reflect
the problems of some of the youngest firms in the industry who be
lieve in new materials and methods and have built up higri produc
tion capacities. If prefabrication is to mean revolution, these are the
revolutionaries, and their Association is well versed in the new skills
of securing programs of government assistance for enterprises likely
to be of public benefit.

173

Part J_ X •

Chapter

DESIGN

I. Introduction

Prefabricators of houses in the United States during the period of
study by no means pursued the same goals. Their diversity of inter
ests is reflected in their approach to design. To some this term meant
structural engineering; to others it meant production engineering; to
a few it meant architecture; and to many it meant sales appeal. The
term properly includes all these aspects, and many others, for a de
cision made in any part of the long operational channel which leads
from raw materials to completed houses may have an important
effect on the design of the house itself.

Considering the term as broadly as this, one might with some
justification say that this entire book is a discussion of factors which
should influence design. As used in this chapter, however, the word
means something narrower and more concrete. Described here in
some detail are the different products which were made by the com
panies studied, with some reference to the techniques by which they
were made. This, then, is design, in the terms of plans and specifica
tions, and as defined by production systems.

That the subject does not lend itself to simple treatment can be
illustrated on the one hand by the millions of dollars spent by
Lustron before even starting production, and on the other hand by
the small company which, in answer to our request for information,
reported that it had been so busy getting into production that it had
had no time to make plans and specifications.

In large part, differences in design stemmed from differences in
basic approach to prefabrication. The type of market sought, the
house planned for that market, the scheme for the production of that
house— all these things varied tremendously, and it would be a fasci
nating study to analyze the reasons of background, experience, intui
tion, and prejudice which could lead to such differences among pro
ducers in the same general field.

One generalization may safely be made, however: the fundamental
decisions upon which these different schemes were based were rarely
the result of a thorough investigation of the whole problem; they
did not come as the result of careful research. Whether research
had a separate existence or was in effect just another of the responsi
bilities of the top management, its scope seems to have been largely

177

limited to the improvement of detail, the saving of material, and the
speeding of operations.

The average prefabricator seemed to think a great deal harder about
the details of his design after it had been adopted and was going into
the production or even the marketing phase than he had in the first
place about the broad principles upon which the design was based.
This is perhaps understandable, since time, energy, and money for
broad analysis are often very limited once operations are under way,
while in the early stages of organization and design the problems of
financing and of creating a production and distribution system tend
to seem very small and remote. When the realization comes that the
first step in the pattern of operations should have been altered in
order better to perform the last, the die has been cast.

In many cases, the prefabricators set out to do little more than
produce a conventional wood frame house by somewhat different
methods and for about the same price, the new methods being under
taken solely from the point of view of reducing production costs. De
sign efforts were concentrated on the selection of materials, of fabri
cation procedures, and of packaging, shipping, and erection tech
niques. In time, and with a large enough volume of business, such
producers might hope to sell a better house for less money. At pres
ent they would say more often that they are selling a better value for
the same money.

A larger group have set out to simplify the design as well as the
construction of this conventional house, so that it might be easier to
build, ship, and erect, and at least as good. Frequently these pre
fabricators have attempted to improve the space arrangements, the
details, the appearance, and the general architectural design of the
houses they build. But they have not usually moved in this direction
beyond their ideas of current public acceptance, or perhaps beyond
their interpretation of the ideas of public acceptance currently held
by mortgage bankers. The industry well knows that it sells its
houses to bankers rather than to purchasers; broad circulation
is given to reports of companies which have brought out houses of
radically simplified or of purportedly modern design, only to fail or
lose money as a result. There have been such cases, and some justi
fication exists for the feeling that good modern design does not carry
with it the strong sales appeal that the predictions of the war years
had attributed to it. In several cases x prefabricators were forced to
stop production on models which had been given much favorable

1 For example: Shelter Industries, Green's Ready-Built.

178

comment in architectural magazines in favor of models of far more
conventional appearance.

Among the prefabricators there were a few, as there have been all
through the years, who approached the problem with a real determina
tion to seek out the basic facts of housing design and to strike out be
yond the limitations of conventional methods. Of these pioneers,
some strove for what has been called in England "austere" shelter:
smaller houses, simpler in plan and construction, of less expensive
materials, and more highly organized in their various functions than
the conventional house ordinarily thought of as "minimum." These
schemes were based upon a desire to find some sort of decent shelter
which might be made available to a wide range of low-income fami
lies. There were also schemes based upon emergency conditions and
designed for temporary or at most periodic use; schemes of this sort
were often designed for use in war production areas. Still others re
sulted from the attempt to achieve a high degree of mobility, with
the consequent desire to cut the weight and bulk to be moved (a few
turned their attention frankly to the problems of the house trailer),
or from the desire to capitalize on the possibilities of obtaining a high
degree of elasticity by means of a very standardized production
system.

Some of these departures from the conventional were very radical
indeed, based on the theory that true mass production will eventually
have to make use of metals rather than wood, aimed at the exploita
tion of some new use of metals or other materials, or guided by the
determined effort to rationalize the whole structural theory of mass
production of houses.

The most familiar example of such a pioneering approach was Buck-
minster Fuller's hemispherical aluminum house, a structure of true
stressed skin design making extensive use of metals in tension rather
than in compression, although as a production, erection, and sales
proposition it was perhaps foredoomed to failure. From an entirely
different point of view unconventional design principles were ex
plored through the work of Wallace Neff, whose gunite structures
were built up over balloon forms, and of R. G. LeTourneau, whose
gigantic traveling forms were capable of carrying complete concrete
houses, poured in one operation, and placing them at the selected
site.

The similarity to conventional construction stood out more than
any degree of innovation, however. This is not necessarily a criticism,
for construction has moved forward, and the conventional house of

179

today has many features of design and construction which differ from
those of the conventional house of only a few years ago.

II. Classification of Prefabrication Systems

Methods of classifying prefabricated houses are as varied as the
purposes of those making the classifications. The general public is
probably most interested in size and price, but these are also the
most variable of characteristics and the least suited to broad analysis.
Architectural style is perhaps the next mest popular basis of classi
fication, and it will be discussed briefly later; attention will also be
devoted to classification by structural system. First of all, however,
attention is given to classification by the principal materials used in
the house, since this offers the opportunity for a brief description of
the characteristic qualities of the various materials for prefabrication
purposes and thus provides a general background for the systematic
analysis which follows.

A. By Materials

Materials have been chosen for ease of procurement and use, for
adaptability to the prefabricated pattern of operations, and for tech
nical satisfaction of normal performance requirements, the special
qualities required in materials by most prefabricators being light
weight, strength, wearing quality, adaptability to normal fabrication
and transportation methods, and as low cost as possible.

1. Wood Lumber and Plywood

By far the largest group of prefabricators at the time of the survey
used wood as the principal structural material. Of the companies
studied, 92 used wood, and, of these, 61 used plywood.

180

The development of plywood construction systems by the U. S.
Forest Products Laboratory and others had, in fact, a large influence
on the growth of prefabrication as a whole. The material is very
light and strong, is extremely stiff, has some insulation value, comes
in large sheets readily adapted to mass-production uses, has fairly
good dimensional stability, and is reasonably durable and low in
cost. It can be used to combine several different functions; for ex
ample, a single sheet can be both surface and structural sheathing on
the outside, or both surface and wallboard on the inside.

Nearly all the companies using plywood used Douglas Fir rotary-
cut veneer, although a few used other types, such as gumwood and
yellow pine plywood, edge-grain fir panels for flooring, or oak ply
wood for flooring. Some of the problems involved in using this
material are discussed in Chapter 9. Its steady rise in cost has been
discouraging to many manufacturers, but most of them felt that it was
the best material available for their needs.

Wood lumber, traditional material for domestic construction in
most parts of the United States, enjoyed great popularity among the
prefabricators, particularly because of its wide public acceptance and
the long experience of builders in making houses of it. From the
point of view of design, members fabricated from wood generally
have the necessary strength, rigidity, and thermal-insulation value at
a suitable cost, although the material offers certain complications in
the factory (discussed in Chapter 9). Wood lumber, in other than
shop grades, was in fairly good supply at the time of the survey, and
its initial cost was low compared to that of other materials. Its char
acter as a handicraft material was actually desirable in the opinion
of most prefabricators, who dealt with a few houses at a time rather
than mass production, and who had frequent occasion to change
shapes and sizes to fit evolving needs.2

2. Steel

Steel is the basic manufacturing material of United States industry,
and there have been many attempts to use it for the manufacture of
low-cost housing. Of the companies in the survey, 13 used steel as a

2 For a very complete discussion of the use of wood and plywood for this pur
pose, see Manual on Wood Construction for Prefabricated Houses, prepared by
the Forest Products Laboratory in collaboration with HHFA (Washington, 1947).

181

basic material in their structure. Of these, three used steel in com
bination with wood and three in combination with aluminum.

In housing design, steel has many disadvantages to counter its
known advantages. Its thermal conductivity is more than 300 times
that of wood so that careful attention must be paid to problems of
heat loss and condensation. Its tendency to rust means that it must
be carefully protected from contact with oxidizing atmospheres, and
this raises costs. Its uniformity and strength are very high but diffi
cult to exploit to the fullest degree, so that much steel is often wasted
in overdesign. Occasionally, there is further waste in pointless imi
tation of wood design. Further, its use requires special attention to
problems of sound transmission and reflection. Nevertheless the
cost of steel and its adaptability to manufacturing techniques will
doubtless continue to appeal to designers. In special forms, such as
the porcelain enameled steel used by Higgins and Lustron, it may
have a new order of general sales appeal as well as improved physical
properties.

In the last two decades, despite a great deal of experimentation
with different steel designs in this country, there has not been so
wide an experience with actual fabrication and use as in England.
Recently, however, there is an increasing tendency among even the
more conventional prefabricators to use steel for members in hori
zontal position which carry loads over fairly large spans, such as
floor joists.

3. Aluminum

Of the companies in the survey, 10 made use of aluminum as a
major structural material, either as framing or as exterior structural
covering. Many others were interested in the possibilities of its use
because the expansion of aluminum production facilities during the
war period had given hope of abundant supply, particularly of sheet
aluminum such as is used in aircraft, and of a lowering of price.

Aluminum has some of the disadvantages of steel, including a par
ticularly high thermal conductivity, but it has certain advantages for
housing purposes, including a positive value as reflective insulation
and a strong resistance to serious corrosion under normal atmospheric
conditions. Although it can be welded only with some difficulty and
must be formed with careful attention to its properties, aluminum is
suited to many industrial techniques. More expensive than steel,

182

pound for pound, it is often competitive with that metal when prop
erly designed, processed, put in place, and protected. Care must be
taken in the use of aluminum, however, because electrolytic action
takes place between it and steel, and because it is subject to attack by
free lime in concrete.

As in the case of steel, although there has been a great deal of ex
perimentation in the United States with aluminum construction, and
although the aluminum manufacturers are marketing an increasing
number of products for use in houses, the largest production ex
perience with aluminum houses has been in England, where the air
craft industry has been producing well-designed units in quantity
since the war.

4. Concrete

Generally speaking, prefabricators consider wet-process materials
unsuited to mass-production methods, although there are exceptions,
as when such special fabrication machines as the Tournalayer are used,
in which case the production interest centers in the machine rather
than the houses. Yet, of the companies in the survey, 10 used con
crete as a major structural material, eight of these using it in the
form of precast concrete slabs. Concrete may have distinct advantage
over other materials in original materials cost, but its disadvantages
of weight, bulk, and frangibility have limited its use primarily to
group erections close to the production point of the slabs. In this
country, where wood and steel are still available at relatively low
cost, concrete construction has been by no means so widely studied
and so carefully utilized as in countries where other materials are
almost out of the question for housing.

In recent years lightweight aggregates and foamed concretes have
become increasingly important, since they lighten the slabs and im
prove the otherwise poor thermal-insulation qualities of concrete. A
great deal of effort has also been expended to improve physical quali
ties and speed up the production cycle by steam curing and vacuum
processes, and to reduce the expense of mixing, pouring, and forming
equipment in relation to the quantity of production achieved. Pre-
stressed concrete shows promise of achieving two or three times the
strength of ordinary concrete with the same weight of material and is
being more generally used in the construction industry, but all these

183

processes are still relatively strange to the single-family-house market,
with which the prefabricators were almost exclusively concerned.

5. Plastics

The literal meaning of this word is broad enough to include many
substances not ordinarily thought of as plastics, such as concrete,
brick, and glass. Plastics in the common sense of the word, however,
were used as a major structural material by none of the companies
surveyed. In part, this was undoubtedly the result of the high cost
of most plastics, now and in the foreseeable future, but there is also
evidence that the structural properties of most plastics are inferior to
those of wood, steel, concrete, and aluminum.

The comparatively low modulus of elasticity of most plastics, in
conjunction with fairly high strength, also means that it would be
inefficient to use such a material as a structural member designed
to carry loads, since the amount of plastic required from the point
of view of strength would be far less than the amount necessary to
prevent objectionable deformation due to low modulus of elasticity.

Laminated phenolics, the most seriously considered of the plastics
as a structural material, have three or four times greater strength in
compression than in tension, which makes it difficult to justify using
so expensive a material in tension as a working skin. Where a mate
rial is used as compression, or load-bearing, members, the aim of the
designer usually is to produce members as light and as stiff as possible.
Yet, to achieve the same resistance to deflection in a laminated phenolic
as in a steel compression member, nearly two and a half times the
weight of the steel would be required.

The plastics industry is at work on these problems, and such new
materials as glass-fiber-reinforced polyesters show promise. The
major structural use for plastics, however, remains in the bonding
of plywood and other built-up structural materials.

6. Paper

At the time of the survey several companies were planning the
production of a house designed to use the surprising strength of

184

plastic-impregnated sheets of paper, so formed and glued (either as
a honeycomb or as a series of corrugated layers) as to form a struc
tural core for stressed skin panels of which the skin might be plywood,
aluminum, steel, or possibly paper itself. However, there was no
actual production of houses made of this material.3

B. By Structural System

Those interested in the production aspects of housing have a major
interest in structural systems, but they have shown in the past a com
mon tendency to classify entire structural systems, and particularly
prefabricated house systems, according to the design of the cross sec
tion of the wall. This practice may be very misleading, since the
system employed in the wall is frequently entirely different from that
used elsewhere in the house. For that reason, each prefabricated
house is here broken down into more or less common component
structural parts and the data are classified according to the system
used in these component parts. The designations of these classes,
furthermore, have been carefully selected to bring out production
differences. Thus, while the phrases "frame assembly" and "frame
panel" may indicate the same thing in final structural result, the
difference between them lies in different amounts of factory prefabri-
cation and different procedures for site assembly. Also, a prefabri-
cator producing a conventional house by fabricating room-size panels
is here distinguished from one producing the same house by merely
precutting the various pieces. The basic pattern of operations of the
prefabricator is indicated as well as the final structural scheme.

1. Frame Assembly

The typical frame assembly is the conventional wood frame house,
in the course of construction of which individual framing members
are erected at the site and various insulating and finishing materials
then applied. The precut lumber house is an example of a fabrica
tion system embodying frame assembly principles.

3 More information about these designs is given on pp. 233 ff. A fuller descrip
tion of the material is contained in Chapter 9.

185

Stud

Diagonal

bracing

Diagonal %
sheathing

Building^
paper

Siding
Sill

Figure 14. Conventional Framing Illustrating Construction Terminology

2. Frame Panels

In this classification, the structural members are preassembled in
the form of panels, and some or all of the insulating and finishing
materials are usually applied in the shop in order to save time at the
site. The wall panel produced by what may be called the typical pre-
fabricator is a frame panel, made up of framing lumber with wood
sheathing nailed to it.

3. Stressed Skin Panels

Where the panels are so designed and assembled that the surfacing
elements contribute in a major way to the structural performance of
the whole, the result has been classified as a stressed skin panel.
Typical design of such a panel is described on p. 228. In some cases a
stressed skin action is partially obtained by the use of a single surface
material bonded securely to the structural framing and by this means
developing some stiffness and strength at the contiguous surfaces.
Most constructions having a single factory-applied surface are not
securely enough bonded to develop this added strength, however,
and so are classified in this report as frame panels, rather than
stressed skin panels.

There have been some attempts to approach a monocoque system
of construction, but requirements for openings and difficulties with
internal shapes discourage the development for housing of a true
single prestressed shell. Yet the Harman house made use of the
tension stressed steel sheet construction developed by the Lindsay
Corporation for truck bodies, and even closer approximations were
made in the hemispherical aluminum Fuller house and in NefFs
hemispherical and double paraboloid "balloon house."

It is also true that, to a minor degree, nearly all so-called frame
systems actually place some reliance upon stressed skin principles,
but they are rarely used deliberately to reduce the amount and
weight of the materials used. At least in theory, true stressed skin
design has a better chance of realization by means of the continuous
sheet surface areas, which are well adapted to mass-production in
dustrial processes.

187

4. Solid Panels

The best example of the solid panel is the precast concrete slab,
which is essentially homogeneous throughout. If the amount of rein
forcing steel or the emphasis on such steel in the design is consider
able, or if laminated panels with plywood or asbestos cement facings
are involved, where major structural resistance to load is channeled
into skin or reinforcement, the panels should perhaps logically be
placed under other structural systems. But where the panel is fabri
cated as a solid entity (thus excluding honeycomb core materials),
and where all parts of the panel assume major structural roles, the
designation of solid panels has been used.

5. Poured at Site

This classification includes essentially monolithic structures in
which the emphasis in the prefabrication system tends to fall as much
upon the pouring and forming machinery as upon the house itself.
An interesting monolithic house, poured near the site, was the Le-
Tourneau house, formed in a tremendous and fully mobile perma
nent form known as the Tournalayer.4

This device, and other devices specially designed to make concrete
pouring and forming operations at the site efficient and economical,
have recently been regarded with a great deal of interest in this
country. A factor in this interest has been the recent rapid expansion
of the development of lightweight concretes, offering easier handling,
better surface qualities, far better thermal properties, and a faster
casting cycle than the regular concretes, while retaining sufficient
strength to be self-supporting and avoid the necessity of added fram
ing or skins.

C. Miscellaneous Classifications

Before turning to the question of architectural design, attention
should be given to two aspects of structural design which are im
portant enough to warrant treatment as separate classifications.

4 Further detail is given in Chapter 9.

188

1. Sectional Assembly

The important feature of this classification is not the structural
system or the materials used, but rather the degree to which the house
is preassembled, by panels or otherwise, into complete volume-en
closing units or sections of the final house. Although the Tennessee
Valley Authority was not the first to use the system,5 this is fre
quently referred to as the TVA style of construction because of the
extent to which sectional and truckable houses were used to provide
living quarters for the crews engaged in the various construction
projects in the Tennessee Valley. The houses were easy to trans
port and to put together, and they required a minimum of labor and
confusion at the site, thus freeing facilities and roads for the larger
jobs at hand. These advantages also frequently recommended sec
tional house types to those planning special communities for the pro
duction of war materials during the recent war, though the TVA
houses featured mobility to a degree greater than that required for
most residential areas.

Because design decisions were made by the TVA rather than by
lending institutions, mortgage insuring institutions, or the ultimate
consumers, the result was that the whole construction operation could
be planned with assurance from the start, and a greater degree of
final finish and building in of furniture was provided than might
otherwise be considered a safe risk.

Several different companies produced these houses for the TVA,
and houses with similar design principles, such as the Prenco house
of the Prefabrication Engineering Co. and the house of Prefabricated
Products Co., Inc., were tried out in other parts of the country. The
conditions of normal business, however, are different from those
faced by the TVA, and the obstacles to be overcome are considerably
greater. Nevertheless the TVA experience points up the fact that,
under certain conditions, prefabrication in whole house sections can
do a clearly superior job.

The sectional house of Reliance Homes, Inc., was an example of a
further extension of the TVA principle (see Figure 22). The Reli
ance house was of steel frame construction, faced with corrugated
aluminum over Homasote on the exterior and wallpapered Homasote
on the interior. The house was factory assembled into seven three-
dimensional room-sized sections, which were completely finished with

5 General Housing Corporation's sectional house is described in Part I, Chapter

2, pp. 37-8, footnote 55.

189

wallpaper, wiring, floor covering, kitchen equipment, heating equip
ment, etc. The sections were transported to the site on three trucks,
unloaded by a crane, and assembled into a house in less than a day.6

The AIROH house (see Figure 23) in Great Britain, manufactured
for the government by Aircraft Industry Research on Housing, is an
example of the use of these general techniques in a light aluminum
house which has been mass produced in tremendous volume for
general residential purposes.

The sectional idea has been carried even further when the complete
house has been made available in one piece, as in the case of the well-
known house trailer. Several designers have taken this trailer con
cept and expanded it by ingenious means to produce a true prefabri
cated house in a single section. Perhaps the best example of this is
the Wingfoot Home of the Goodyear Tire & Rubber Co. (see Figure
26). Similar in concept to the earlier Stout Folding House,7 it was a
fully preassembled and prefinished flat-roofed house of stressed skin
plywood with utilities completely installed and ready for connection
to municipal services. There were two bedroom sections which
could be pulled out of the central section in the manner of drawers,
expanding the original trailer to a living unit of 253 sq. ft.8

These were houses of which it is fair to say that the makers' aim
was not so much to supply permanent living quarters for a complete
family as to provide temporary houses with far more livability than
the normal trailer. Another house under design during the period
of the survey, however, carried the same expanding principle even
further to produce a good-sized permanent house. This was the
house designed by Acorn Houses, Inc. (see Figure 27) to move over
the roads in a low trailer bed and unfold at the site into a two-bed
room house of 800 sq. ft. This was made possible by the use of walls,
floors, and flat ceiling-roofs of plastic-impregnated paper core with
bonded plywood skin, having a cross section thin enough to permit
the folding of hinged walls, floors, and roofs against the central
utility core during transit. At the site, girders were laid on posts, the
floor units unfolded downwards, the walls unfolded outwards, and the
roof unfolded over the whole and bolted down. The scheme per-

6 Reliance designs were considerably changed in the period following the sur
vey. The latest schemes divide 'the house into three sections, one nesting within
the other, so that an entire house can be carried on a single trailer.

7 Developed in 1937, this was a fully mobile trailer which could be expanded
to about three times its original size by folding the side walls up and out to form
an additional room on either side.

8 Construction of ceiling is described on p. 251.

190

mitted complete factory finishing with a minimum risk of damage in
transit or in construction.9

2. Modular Design

A great deal of emphasis has been placed upon the principle of the
dimensional coordination of building materials and components.
This principle must be distinguished from the so-called modular
planning long used by many architects in working out plans, although
the two have elements in common. Architects' modules have been
for the most part space planning tools, used as a means of assuring
repetitive structure and planning simplicity at a larger scale, the
modules for such purposes running from 3' or 4' in the case of houses
to 20' or more in the case of office buildings and factories.

An illustration of the use of architectural planning of this sort may
be found in General Panel, which was originally designed to sell not
houses at all, but only structural panels in modular sizes (based on a
module of 40" ) and varying styles, capable of assembly into an infinite
variety of houses or other buildings in accordance with the demands
of the individual consumer (see Figure 15). It has long been the
feeling of Walter Gropius, one of the original developers of the Gen
eral Panel system, that the best combination of mass-production effi
ciency and of marketing flexibility could in this way be achieved. At
the time of the survey, however, a relatively small percentage of the
business of General Panel Corporation of California had been along
these lines.10 A few other prefabricators made additional business
for themselves by selling their panels for incorporation by local archi
tects and builders into houses of conventional construction. The,
HomeOla Corporation sold separate panels several times, and some
of the resulting houses were given acclaim in the architectural mag
azines.

Modular design is of obvious importance in prefabrication. Few
prefabricators, however, have understood it as a basic principle of de-

9 Further details on construction are given on p. 235.

10 For many reasons, it has been necessary for prefabricators to concentrate on
producing a complete house. These modular panels have not been generally
available, although recent efforts have been made to bring them into more gen
eral use. Even when offering packages of panels to be assembled into specific
designs, General Panel has been able to take advantage of this basic flexibility
and to offer as many as 25 radically different designs.

191

THE JOINT
is assembled when
three parts are
nested together and
the fourth driven
home with a hammer.
The system is
designed to permit
joining of panels in
any combination

Insulation
Bracing

Exterior

skin

Wiring

Post

Corner

One way

Two way

Three way

Four way

THE PANELS
are interchangeable.
The entire house is
panelized, all panels
being similar in
proportion, edge
profile and method
of connection

Figure 15. The General Panel System

sign, applicable to all parts of the structure. These basic principles
have been briefly sketched 11 as they were developed by Albert
Far well Bemis. All dimensions in Bemis' experimental prefabricated
houses (1925-1932) were based on his cubical modular method of
design, including dimensions of finish materials and of some equip
ment. He demonstrated that one result was complete flexibility of
layout and theorized that substantial production economies also
should obtain.

There are at least two factors which explain why prefabricators do
not at present make greater use of modular design. The first is that
they have standardized on only one or two or at least a very limited
number of house plans. With so little variation of design, complete
modularization carries small advantage. In the second place, the
building materials industry has only just begun to standardize dimen
sions on a modular basis, and much of this early coordination has been
accomplished by the brick, tile, and masonry unit manufacturers so
that it is of little use in prefabrication. Consequently, prefabricators
have been forced to choose their materials and equipment from a
poorly coordinated industry, and to design their houses around them
in the most effective manner to meet immediate needs. Such prob
lems as the interchangeability of wood and metal windows, free choice
of built-in mechanical equipment, and complete flexibility of layout
have been left to the future.

Although none of the companies in our survey had completely
adopted the modular theory of design, a majority of them were bene
fiting by the use of some planning or manufacturing module. The
most common such module was 4", or some multiple thereof. ( Amer
ican Standards Association official American Standard No. A62. 1-1945
states, "The basis for dimensional coordination shall be the standard
grid based on the module of 4 inches.")

At least 46 of the companies dimensioned their components in
multiples of 2" or 4" and, of these, six favored the 40" manufacturing
module, which is not only practicable for local purposes but also is
close to the metric module widely favored abroad. Some sort of
recognition of the 4" module was given by at least 62 companies in
all.

Other modules were used, however. At least three companies
used a basic module of 3"; two used 39" as a manufacturing module;
and one used 4' 3%" as a manufacturing module.

11 Part I, Chapter 2. Extensive design details are available in the A62 Guide
for Modular Coordination (Boston: Modular Service Association, 1946).

193

Of those not using modules as a design or manufacturing basis, most
were producing but one or a very few standardized models which
permitted a relatively standardized production. Some claimed that,
by disregarding modular dimensions, they were able in practice to
effect saving in the fitting of equipment and the sizing of rooms. On
the other hand, others making a limited number of models still
found advantages in modular dimensioning. Thus Harnischfeger
made up panels in widths of any multiple of 4', and found that it gave
his dealers a good deal of elasticity in the erection procedures selected.
Some preferred to handle 4' panels as such; others asked to have them
preassembled in 12' and 16' lengths; and at least one asked for de
livery in the form of fully assembled wall-length panels, to be erected
at the site with a crane.

3. Architectural Style

One of the oldest and certainly one of the most popular methods
of classifying houses is by the general appearance or architectural
style. Aside from simply describing the basic surface material, this
is indeed probably the way most people try to describe houses to
one another. Those trained in architecture, however, would be the
first to say that this is something less than the ideal method, because
general appearance and architectural style may mean very different
things to different people.

As has been said at the beginning of this chapter, this entire book
is concerned with architecture in the broad sense. The materials
and structural design, the production system, and the erection
scheme are far more important to the architecture of prefabrication
than the so-called style in which the final house is clothed. Yet classi
fication by architectural style cannot be entirely dismissed here, be
cause it is a matter of great concern to most prefabricators, however
widely their interpretations of architectural treatment may vary.

A few prefabricators, including some of the prominent names in the
business, have little use for architects and profess to believe they have
nothing to offer the prefabricator. At the other extreme, several pre
fabricators have come into the field directly from architecture. Some
of these are crusaders, and a few appear to be far more interested in
expanding the vision of the public and of their profession than in
making houses on a business basis.

194

For most of the prefabricators, however, good architecture is but
one of several important aspects of the business, and they take steps
to get it, according to their understanding of the term. As might be
expected, the most rational architectural approach has generally been
found in the companies making the greatest innovations, for in such
cases the importance of the architect in producing a livable and sal
able house by the new techniques becomes obvious. Probably archi
tecture has had the least influence on those companies devoted to
manufacturing, by a somewhat more industrialized process, the same
kind of conventional house as is built in the area by speculative build
ers. Yet even these often found a business advantage in devoting
time and money to the appearance of the house and to the gadgets
and decorations which frequently pass for architecture. The adver
tising world has created too vivid a picture of the normal American
dream house to be disregarded.

In factual terms, by far the majority of the prefabricators during
the period of survey were satisfied to put before the public what can
only be called conventional houses, either as the result of careful de
liberation or only unconsciously, because that was what a house had
always meant to them. At least 80 companies came in that group.
On the other hand, perhaps 40 wanted something different, usually
along the lines of what has been called "modern" architecture. Archi
tects would probably classify as of sound design— whether conven
tional or modern in spirit— about one-quarter of these prefabricated
houses, or very little more than would be the case with conven
tional houses. In about the same fraction of houses, often although
not always the same, were the services of an architect employed some
where along the line.

The value of architecture to most companies lay in its relation to
marketing, and it was for the most part thought of primarily as a sales
feature. Some prefabricators, in fact, spoke of a "basic house" to
which such "architectural treatment" as false gables, long shutters,
and special entrance details were to be added, often as extras. They
were usually convinced that houses of modest and conservative ap
pearance, reminiscent of the Cape Cod cottage, represented the safest
gamble, and in this opinion they were reinforced by the commercial
failure of several attempts to market more advanced designs. Many
were, however, becoming convinced that in recent years the archi
tectural tastes of the public have in some respects been tending to
move away from the Cape Cod cottage. The suggestions of open
planning contained in such terms as "picture window" and "ranch-
house style" were becoming stronger, the more so the farther west in

195

the country, and the wide appeal of the all-on-one-floor house was
recognized. Few, if any, of the larger companies produced a two-
story house. A sizable group produced, and perhaps twice as many
were contemplating, a story-and-a-half house, usually of conven
tional design with two bedrooms planned for the second floor but not
finished, but in most cases this was done primarily for reasons of
economy.

The largest single factor in making the prefabricators conscious of a
more fundamental sort of architecture in terms of sound space plan
ning and construction probably was the requirement, often a matter
of life or death, that their houses meet the approval of the FHA. A
certain minimum good design was assured in this way, but it is
clear that the narrow views of many financiers on architectural mat
ters were a severe limitation on those whose training was good and
who were eager to offer an architecturally sound house created by a
new approach.

The Reliance house, first designed by William Lescaze with a flat
roof, was refused approval for mortgage insurance by the Philadel
phia FHA office, though the design met with no objections from the
national FHA office. The Philadelphia office, stating that the design
lacked "to a substantial degree those essential esthetic qualities and
visual appeal which are necessary to assure continued marketabil
ity/'12 required that a pitched roof be added to qualify the house
for mortgage insurance. The local office later did permit the flat-
roofed houses to be erected, after finding that their acceptability ex
ceeded that of the pitched-roof variety.

III. Description of Components

A. General

In the discussion which follows, the system of classification by struc
tural system is applied in detail, not to the house, but to its major
components, using for this purpose those components into which the

Architectural Forum, 88 (March 1948), 11.
196

Chart A

Roof

Ceiling

Floor

Frame Assembly

! ! ! ! ! I Frame Panel

Stressed Skin Panel

Solid Panel

Construction Used in
Prefabricated Components

Notes: Design information was analyzed for
125 companies in all. In some cases the
information required to prepare these charts
was not available. In other cases, companies
could properly be listed in more than one
category. The totals should not, therefore, be
expected to check with the headings in every
case. At least 25 companies used concrete
slab floors for most of their houses. Many
more used slab floors occasionally.

house is most readily divided, namely: foundations, floors, walls, ceil
ings, and roofs. Chart A gives a breakdown of companies according
to the structural system used in their various component parts.

For each component, there is a breakdown according to structural
system (as defined in the last section), and within each such subdi
vision, as fully detailed information is given as possible. Thus, in
order to find the wall-panel size most commonly used by prefabri-
cators of houses of stressed skin plywood construction, it is necessary
to turn to the component "walls," and under it to the structural classi
fication "stressed skin." (In this case, there was no single preference;
prefabricators were almost evenly divided between 48" panels and
room-size panels. ) So far as possible, this information has been made
available in tabular form, and from a brief scanning of these tables
the characteristic construction systems become apparent.

It should be recalled that the survey did not include the entire
industry, that there was a predominance in numbers of small-shop
fabricators of essentially conventional houses, and that the producers
of potentially great numbers of new types of houses were in few cases
in production and in no case in full production. The numbers which
appear in these tables, therefore, are not suitable for statistical analy
sis; they serve rather to give a generally accurate picture of the
industry.

One general comment which is made here in order to give it due
importance concerns the treatment of detail. In most of the houses
studied, particularly the interiors, there was an element of crudity-
lack of refinement of details, lack of precision of manufacture, and
insufficient attention to materials used for interior finish— which could
well do injury to the whole product. By and large, this was no more
true of prefabricated houses than of conventional houses built during
the same period, but the prefabricators could far less afford to have
criticism focus on such matters.

B. Foundations

Very few house manufacturers supply any sort of prefabricated
foundation, and there are almost as few specialized manufacturers of
prefabricated foundations. Several prefabricators did supply concrete
posts or wooden piers, however, and two companies even had precast
slabs for use in forming basement walls and floors, but except for
specialized local operations prefabricators saw no economies inherent

198

in the use of precast concrete slabs for basements because of their
weight and bulk and general difficulty of handling, and because simple
means of construction at the site were readily available. Neverthe
less, special handling devices make it possible to move precast slabs
into place, and several companies were working at least experiment
ally on the simplification of their foundation construction work.

Of the companies studied in the survey, at least 16 designed their
houses specifically for a basement and supplied detailed foundation
plans, although in most cases the basement was to be built by the
local builder. These prefabricators were predominantly in the north,
where continuous foundation footings to a depth as great as 4' might
be called for anyway, but they also felt that the public wants base
ments, as shown by almost all conventional houses in some areas.
The arguments for the basement stress the large amount of storage
and general utility space thus made available at a relatively low
cost, and these considerations have enough weight to persuade at
least 47 of the other prefabricators, who do not insist upon base
ments, to offer them as an optional feature.

Those opposed to the use of basements point out that they add
cost to the final house— on the average about $500— and that the func
tions usually allocated to them can more safely and efficiently be per
formed above ground in space designed for the purpose. Essentially,
these men say, the basement is incompatible with the concept of
prefabrication which would reduce site work to a minimum, and
which requires the timing of site preparation to be as simple and de
pendable as possible. To schedule steady sales throughout the year,
a northern dealer would have to tie up money in many basements
made ready in good weather to handle house erections in bad.

This argument depends upon many design factors, and cannot be
settled once and for all,13 but the advantages to the prefabricator of
the basementless house are such that there has been a strong tendency
to build such houses, even in northern climates. At least 56 prefabri
cators produced basementless houses exclusively. A few of those ex
perimenting in northern and eastern areas with such houses found
better public acceptance than they had expected, particularly if con
struction economies were passed along to the consumer. On the
other hand, several companies selling large quantities in northern cli
mates when they gave the option of basement or no basement found
that the market preferred the basement. Of course, this may be

13 An interesting study has been made by the HHFA on "Basements vs. No
Basements for Houses," HHFA Technical Bulletin, no. 8 (January 1949), pp.
47-59.

199

attributable to the nature of the differences in the design and cost
of the two types as much as to a preference for the basement as
such.

The majority of basementless houses were placed on continuous
foundation walls of some sort— either poured concrete walls support
ing a wooden or concrete floor system, or the edge of a floating slab
designed in effect as a grade beam. Only a few of the houses, 15 in
all, and mostly the smallest and least substantial, were placed on piers
or posts. This was in part the result of FHA, building-code, and bank
ing requirements of continuous foundation walls.14

Little new development has taken place in foundations, and that
has been done mostly in connection with basementless houses. Some
of the ideas developed for houses on pier or post foundations include
the use of special built-in jacks to level the house on the permanent
foundation system (TVA), the use of precast concrete discs with
holes in the middle strung on an iron pipe to build up masonry posts
(Swedish), and the use of precast concrete posts which are hung
from the jacked-up floor beams (see Figure 16) until the bulk con
crete footings poured about the posts have had time to harden suffi
ciently to permit the removal of jacks (Acorn).15

With the grade beam foundation, which is not necessarily footed
below the frost line, there has been quite a bit of experimentation,
and study is in progress. In at least one case it was proposed to use
radiant heat in such a slab to prevent frost formation in the ground
beneath it. Most designs, however, set out to defeat frost heave by
using sand and gravel under the slab and by otherwise naturally
or artificially keeping the underlying soil well drained. Some de
signers claim that the loading of many domestic superstructures is
so light that little damage is likely, to structure, foundation, or
plumbing, as the result of the temporary lifting of a corner through
mild frost heave.

Perhaps the second largest problem of the grade beam or slab
foundation is that of insulating the walls and the corners and edges
of the floor from the cold. This problem has been the subject of a
great deal of study and is more fully discussed on p. 208.

14 These requirements are becoming more liberal, and a pier or post founda
tion, with proper insulation in the floor, is now acceptable in many more localities
than at the time of the survey.

15 Recent work by HHFA engineers tends to show that adequate ventilation of
the crawl space under such houses is the only surely effective means of preventing
accumulation of moisture in the wood of the structure above (HHFA Technical
Bulletin, no. 8 [January 1949], p. 107).

200

STEP1

Girder bolted to precast concrete
pier and lowered by gin pole into
hole in ground. Jacks used to
support girder and pier

STEP 2

Girder is leveled by means
of two jacks. Wood blocking
inserted to support girder
and pier. Jacks removed

STEPS

Concrete poured into
hole and allowed to
set Hole is then filled
with earth and well
tamped

Figure 16. The Acorn Footing

C. Floors

1. Frame Assembly

Under this heading are included floor structures which are made
up at the site from precut or otherwise prepared members. A large

Chart B

Frame Assembly Floor
32 Companies

Frame members

Structural floor

wood

steel

2" X 8"

2" X 6"

16" o.c.

nailed

bolted

plywood

wood

nailed

bolted

prefinished hardwood

unfinished hardwood

nailed

glued

glued and nailed

Finish floor

number of manufacturers who panelized walls and other portions
of their houses preferred to precut the floor system, probably be
cause of the bulkiness of panels built up with joist-size lumber, the
difficulty of achieving a solid and silent structure, the known mar-

202

ket preference for a continuous hardwood floor, and the relative
advantages of site assembly over prefabrication in making allowance
for imperfectly dimensioned foundations and in permitting the use
of diagonal lumber subflooring. In all, 29 companies used frame
assemblies of wood in their floors.

It is significant, however, that three of the largest companies
(National Homes, HomeOla, and Houston Ready-Cut) made use
of wide-spaced (4' o.c.) steel framing members (I beams, channels,
or open web bar joists) on which framed wood sections were placed,
and at least as many more were considering use of this system.

A summary of the data on frame assembly floors is presented here
with (Chart B). This chart and the others which follow it give
only a selection of the most useful information from our survey.

Although no such system was actually seen in use in a prefabri
cated house, the National Lumber Manufacturers Association has
publicized the design of a floor of 2" X 6" dressed and matched
tongue and groove planks laid over girders 6' 0" o.c. According to
the preliminary figures of the Association, this offered hope of sav
ings as high as 26% on labor and 14% on material in comparison with
the conventional system, and might add as much as 24% to the in
sulation value. It had been used by architects and was being seri
ously considered by several of the prefabricators.

2. Frame Panels

Chart C presents a summary of the details of construction used by
the 49 companies which employed frame panels in their floors. The
variation in sizes and spacings of floor joists in these panels is the
result not so much of building regulations or differences in engi
neering standards as of variation in structural floor (usually sub-
floor) thickness and design.

Floor panels used in basementless houses are usually insulated,
particularly in northern climates, and the importance of vapor bar
riers is beginning to be realized. For the installation of insulation
and vapor barriers, factory assembly appeared to offer some advan
tages over field installation, although only 14 companies gave definite
indication of providing both, and among them there was wide varia
tion, both in materials used and in method of installation. Some
took the chance that insulation, even thin reflective insulation, might
be damaged in transportation and handling, and made no effort to

203

Chart C

Frame Panel Floor
49 Companies

Frame members

Panel size

Joint

Structural floor

Finish floor

* Types discussed pp. 218 ff.

wood

steel

2"X%"

2" X 6"

2" X3"

16" o.c.

24" o.c.

nailed

glued and

nailed

4' X 12'

8' X 12'

6' X 12'

butt

interlock

spline

wood

plywood

steel

metal lath

and concrete

«!

nailed

glued and nailed

prefinished

hardwood

unfinished

hardwood

nailed

glued

poured

204

protect it or box it in, probably on the theory that where boxing can
be avoided and bridging done at the site there are certain advantages
of nesting panels during shipment.

The difficulties of factory application of finish flooring are empha
sized by the relatively small number of companies which attempt
it. Unless joints 16 between panels, always a problem in floors, can
be concealed under partitions and thresholds, it becomes a handi
craft operation to make them tight, and there are additional problems
of protecting the assembled and sanded floor in handling. Since
floor panels also tend to be large, the factory application of finish
flooring tends to make them hard to manhandle, particularly if
lumber subflooring is used. With rougher floor panels, simple butt
or lap joints can be used, and the finish floor applied in the field.

In order to avoid doubling of framing members at the joint, sev
eral companies have changed from butt joints to some variation of
lap joints, which require a higher degree of subfloor uniformity and
thus place some advantage on the use of plywood for the structural
floor instead of lumber.

3. Stressed Skin Panels

Chart D presents a summary of the construction details of the 16
companies using stressed skin panels in their floors. From the point
of view of reducing weight, stressed skin panels have definite advan
tages, and the double skin makes a substantial increase in the
thermal-insulation value of the floor. It further obviates any neces
sity for bridging between joists. However, some care is required to
prevent condensation within the panels.

It will be noted that these panels are not produced in noticeably
larger sizes than the frame panels, despite the possibilities offered
by lighter weight, and that the joints in such panels tend to become
more complex because of the inability to get at the interior of the
panels, although in some cases hand holes were provided as access
for bolting or other inside connection. These joints, in addition to
providing continuous structural connection, were used to position
the panels, and a feature was often made of their special characteris
tic of making the whole structure demountable. In some cases,
however, they become so complex as to require rather expensive
millwork pieces.

16 Joint types are discussed on pp. 218ff.

205

Chart D

Stressed Skin Panel Floor
16 Companies

Frame members

Panel size

Joint

Structural floor

wood

steel

aluminum

2"X6"

1"X6"

2" X 8*

2" X 4"

16" o.c.

glued and

nailed

glued

bolted

4' X 8'

4' X 12'

4' X house width

butt

spline

interlock

lap

plywood

glued and

nailed

glued

electronic

glue

riveted or

bolted

206

Finish floor

plywood (3 rotary, 2 edge grain)

unfinished hardwood

optional

asphalt tile

glued

nailed

glued and nailed

stapled

At the time of the study new interest was being displayed in the
use of edge-grain plywood as a finish floor and subfloor combined,
to be applied in one sheet. This was used by only two companies,
and it was not known at the time whether a cost saving could be
made in this way or not. However, it was generally felt that rotary-
cut fir plywood would not be satisfactory as finish flooring, because
of its relatively poor and uneven wearing quality. One company
used oak veneer as a factory-finished floor surface, but it appeared
to be rather expensive at the time. There was little experimentation
with the new composition floor materials.

Nearly all companies used linoleum flooring in the bathroom or
kitchen or both, but none attempted to apply this in the factory.
It was interesting to note an increasing tendency to use asphalt tile
flooring throughout the house, in the case of wood floor systems as
well as that of concrete, and, when used, it seemed to meet with
little marketing resistance.

4. Solid Panels

Only five companies used solid panels in their floors, and these
were for the most part precast concrete panels. The fact that so
few companies used such floors may be explained in part by the fact
that only one smooth surface is required for a floor, and companies
using precast wall slabs find it simpler to cast the floor slab on the
ground at the site.17

Where there is a basement there is the possibility, not actually
tried by any of the companies in the survey, of constructing a floor

17 Nevertheless, there are indications that certain techniques, such as the
Vacuum Concrete process and the use of prestressed reinforcing, together with
some means of bedding them down firmly, may make it economically possible to
use precast concrete floor slabs, particularly in large projects.

207

with long precast and prestressed reinforced concrete beams which
have a rectangular cross section and usually a hollow central core.
Such beams were widely marketed by The Flexicore Co., Inc., and at
least two other companies for use in conventional construction, and
they offered interesting possibilities for specialized construction if
joint problems could be handled.

5. Poured at Site

At least 25 companies were using poured-at-site floor construction,
the great majority of them using asphalt tile or linoleum for their
finished floor surface.

Since concrete is a porous material and a poor thermal insulator,
it was becoming increasingly obvious to these companies that care
ful attention must be paid to insulating it from both the ground and
the outside air (see Figure 17). Western firms, with longer experience
in this sort of construction, were often found to exercise great care
in the placement of a waterproof membrane beneath and around
the edges of the slab— usually hot tar and #15 roofing felt— even
where no insulation was required.18 Many of the slabs produced for
northern climates, however, seemed deficient in insulation.19 The
possible use of lightweight aggregates having better insulative quali
ties and of waterproofing admixtures was under consideration at the
time of the survey.

The development of the ground slab was being spurred not only
by materials savings, labor simplicity, time savings, and generally
lower costs, but also in some degree by the regulations of the FHA

18 The Byrne Organization takes the following precautions with slabs cast di
rectly on the ground:

"All slabs have a porous sub-base of considerable thickness with a perimeter
grade beam around the building sufficiently deep to be below the frost line and
to cut off subsurface water. . . . We never locate slabs on lots which have an
accumulation of water through poor drainage. The top of the slab is furthermore
placed about a foot above the ground which is carefully graded away from the
building on all four sides." (Letter from Wesley H. Blank, Chief Engineer of
the Byrne Organization, to the Bemis Foundation, July 31, 1947.)

19 See "Insulation of Concrete Floors in Dwellings," HHFA Technical Bulle
tin, no. 8 (January 1949), p. 149. Also Concrete Floors for Basementless Houses,
Small Homes Council, University of Illinois, Circular Series F4.3 (August 10,
1948).

208

requiring a large ventilation space under a platform floor and above
the ground in the case of a basementless house. Many companies
felt that their small houses would have an awkward appearance and

Metal flashing
Asbestos cement
Insulation
Anchor bolt
Concrete grade
beam

Footings not carried
below grade beam

Anchor bolt
Insulation
Foundation wall
Concrete footing

Footings carried
below grade beam

Concrete
slab

Roofing felt
Cinders
Earth

Figure 17.

Roofing felt
Cinders
Earth

Examples of Grade Beam and Concrete
Slab on Grade

serious marketing difficulties if built up too high above the ground
level, and there was reason to believe that the public was not seri
ously opposed either to the basementless house or to the ground slab
floor.

209

D. Walls

1. General

Prefabricators, housing theorists, and rationalizers of construction
in general have devoted more attention to the construction of walls
than to that of any other part of the house. The human being,
viewing the world primarily in a horizontal plane, seems to assume
that walls make the house, and the inventive mind has long dwelled
on the possibilities of creating an ideal material to serve all the
functions of the wall cheaply and efficiently.20

Nearly every prefabricator manufactured some major part of the
house walls, and many manufactured nothing but the walls and
were looking forward to the day when they could get into the
extensive non-residential field as well. Yet to manufacture the walls,
or walls and roof, leaving the bulk of the house to be provided
locally, is to realize only a part of the potential advantages of pre-
fabrication; indeed, because of the number and variety of openings
required, some prefabricators claim that it is more difficult to manu
facture walls than floors, ceilings, or roofs. To most of the companies
in our survey, however, considerations of rationalization for large-
volume production, or of marketing only stock components for
assembly as desired by the local purchaser, were not important.
These companies were out to make a profit by simplifying somewhat,
improving somewhat, or lowering costs somewhat, without substan
tially altering the normal house as it is known and as it has become
acceptable to the public. The system developed for the production
of walls was, in most cases, at the heart of the whole scheme.

2. Frame Assembly

Precut houses. There was a steady production of what was known
to the trade as a "precut" house, in which not only the walls but the
whole structural frame and much of the finishing material were pre
cut and shipped in a single house package. Such a package rarely

20 Walls and partitions represent from 30% to 40% of the total construction cost
of a house if millwork and interior and exterior finishing are included.

210

contained more than the necessary wooden pieces and possibly some
roofing or flooring materials in addition. Although this system may
offer a few specialized advantages (for example, minimum bulk for
export purposes), it is basically a conventional frame structure with
marketing advantages depending on price and convenience rather
than design. In fact, in most cases, the precut houses were clearly
not better designed from the architectural viewpoint than the aver
age conventional house.

At the time of the survey the National Retail Lumber Dealers
Association had offered the "industry-engineered house," based on
modular coordination in wood frame design, efficient precutting, and
a rational assembly system.21 The Peerless Housing Company, Inc.,
was also at work on precut houses of advanced design, making use
of special assemblies such as trusses and girts and millwork of stand
ardized parts to obviate the need of more complicated assembly, and
further simplifying erection operations by procedures designed to
eliminate the chance of confusing the various precut pieces— a chance
which, together with the tendency of the local erection crew readily
to give up the search for required pieces and cut other pieces to fit,
has long been a special difficulty of the precut house.

Frame and curtain wall construction. This usually involved wide-
spaced framing members which wholly support the roof system,
leaving no structural function for the "curtain walls" applied to these
members. In some cases these systems can be termed "exoskeleton"
systems since their framing members are exposed on at least one
side of the wall. As such they have an unconventional and not
unattractive appearance.

Such a system was used in the house designed by the John B.
Pierce Foundation and produced primarily as the Celotex Cemesto
House (see Figure 24). This house had 4" X 4" posts as much as
12' o.c., with the edges of the Cemesto 22 board curtain panels en
closed within the posts themselves, and with the roof load carried
to the posts by built-up plywood girders placed horizontally at the
top of the posts. Above the 4' high tier of Cemesto panels placed
above the floor and its capping lumber member came a second tier
which, because it served no structural function, provided a great
deal of freedom for the location of windows and other openings.
The system was used extensively during the recent war (together

21 The "industry-engineered house" plan was used by the University of Illinois
Small Homes Council for its time study of construction methods, Research Report
on Construction Methods.

22 A sandwich board made up of fiberboard filler with asbestos cement facings.

211

with a vertical type using posts spaced 4' o.c. and eliminating the
special top girder), and the free fenestration and low cost appear
to have overbalanced the special requirements of the system: high
precision millwork in dimensioning the framing members, and care
ful protection of the edges of the Cemesto panel against fracture in
handling and against moisture in use. Many such houses have been
erected since the war by private builders.

One such builder, Modern Standardized Buildings Co., made use
of the 4' spaced post and Cemesto panel system in a variation specially
designed to avoid the costs of select grade lumber, millwork con
struction, and other expensive finishing detail, and to take advan
tage of the properties of protective mastics and paints. The design
was expected to yield cost savings even with small production volume
because of simplified production and erection procedures.

Another frame and curtain wall system of interest was the so-called
"Ratio Structures" house of Wiener, Sert, and Schulz, developed
during the recent war.23 Exterior wooden posts 13' 4" o.c. supported
longitudinal beams and tie beams which in turn supported a series of
curved plywood-covered panels to form a continuous arched roof. A
secondary framing system in the walls, using a spacing module of
3' 4", was filled with solid insulated wall panels, windows, or doors,
as the case might be. Interior partitions in the temporary war
projects which used this system were built up of laminated fiberboard
on wood frames and jointed by the use of plywood splines, and
fiberboard ceilings were hung from the tier beams. This system,
like the Pierce system, offered great elasticity in the design of open
ings, but the secondary framing system tended to be a needless and
somewhat wasteful duplication of the primary roof support system.

Production Line Structures offered a good example of another
frame and curtain wall system of special interest (see Figure 25).
In this system, half frames composed of wooden members and
nailed plywood gussets were brought together at the site to form,
in effect, three-hinged arches 4' o.c.; these were tied together at the
ridge and eaves by longitudinal members and at the lower ends by
plywood panel members. Continuous openings between structural
members above these panels were filled in at the site by wood
awning-type windows or by solid panels, as the location might dic
tate. End walls were light framed and plywood covered, and they
served structurally only as stiffeners. The house was designed for

23 Modern Designs for Prefabricated and Demountable Buildings, Office of
Production Research and Development (Washington, 1944).

212

ready production, in a standard width but in any multiple of 4' in
length, by the use of standard parts. This design was suitable
primarily to warm climates; its contemporary quality may be shown
by the fact that it received first citation in the 1946 Progressive Archi
tecture Awards.

Metal walls. In this group were metal systems ranging from those
which largely imitate wood frame structure to those in which some
element of stressed skin design is employed in order to take advan
tage of the possibilities offered by metal for production of thin,
strong, and standard sheets. In metal structures the line cannot
easily be drawn between frame assembly, frame panels, and stressed
skin panels. Classification of a system often depended upon
whether, at the time of the survey, the framing members went to
the site as separate elements or preassembled into panels, or whether
at that time the manufacturer was shipping out his parts "knocked
down" or had the time and factory space to do a certain amount of
preassembly. In general, if panels of some sort were preassembled,
the system was classed as frame panel, and if great reliance was
placed on stressed skin design, the system was considered under that
heading.

Typical of the steel systems in which metal studs serve simply to
replace wooden studs was that put on the market by Stran-Steel, in
which the stud was specially designed to permit nailing into it.
Nevertheless, the Stran-Steel Arch Rib Homes— an outgrowth of the
wartime "Quonset" huts— were very different from wood design.
They produced a structure of semicircular arch section, enclosing
the house volume with substantially less material and avoiding the
difficulties usually encountered at the juncture between roof and
wall. In this system, corrugated sheet-metal cladding was applied
to the exterior side of steel ribs spaced 2' o.c., with paint, special
protective coating, or insulation used according to the circumstances.
In many cases, roof sheets were raised from the main framing to
permit continuous ventilation under the roof. Windows, doors, and
other openings along the sides of the structure ordinarily were verti
cally framed bay extensions of the structure, often in wood. The
final product was used in many different situations, frequently pro
ducing very interesting variants on the usual themes of domestic
architecture.

In systems of this sort special care has to be taken to avoid con
densation resulting from contact between the highly conductive
metal skin and the frame. The use of horizontal rather than vertical
corrugation helps to reduce the area of continuous contact between

213

these two, and the furring out of interior wall surfaces on wood
battens, the separation of skin from frame by insulating felt or rubber
strip, and the provision of weep holes and drip on the inside of the
wall to permit escape of moisture which accumulates on the under
side of the skin, are examples of other design devices employed to
improve the residential qualities of this sort of construction.

A frame assembly house construction system of aluminum was
offered by the Fox Metal Products Corporation. In this system, the
basic framing member was made up of two 6" channels of 0.064"
thickness bolted back to back to form an I section into which facings
and insulation could be screwed or nailed; the I sections were spaced
2' o.c. Exterior surfacing was 0.040" thickness aluminum sheet,
crimped for stiffness in such a way as to resemble clapboards. The
crimping reduced the area of contact between exterior surface and
frame, and further protection was offered by an insulating strip of
asphalted felt. The interior surface was generally %" Upson24
board, applied over a V blanket of compressible insulation where
required by climate, and held in place by cold-rolled vertical alumi
num batten strips. Partitions were formed of 4" I-section channels
having Upson board on both surfaces. The roof structure was much
the same as the wall, with additional insulation and with a layer of
Upson board placed immediately beneath the exterior surface as
well. No further finishing was required for wall or roof surfaces.
Windows were wood framed casements, placed in the 2' space be
tween studs.25

Another frame assembly house, basically of steel, was that of the
Harman Corporation. This house was one of the earliest to receive a
guaranteed market contract (for 4,200 houses), and it was later widely
publicized by the failure of the company. This house used the
Lindsay trailer body structural system in applying thin sheet steel
(26 gauge on walls, 24 gauge on roof, galvanized) in tension as an
exterior surface over steel wall studs and roof trusses 39" o.c. When
completed at the site, therefore, this became a stressed skin system,
with the skin bracing the entire structure. An interior lining of
insulation and wallboard was furred out from the steel members with
wooden strips, and insulated partitions were made of light steel

24 Upson board is a laminated fiberboard available in room-sized panels.

25 More recently, Fox Metal Products Corporation has supplanted this model
with its Marquette home, which varies from this description in many respects,
particularly in the use of plywood interior finish, of more extensive insulation,
and of a peaked roof finished with sheathing lumber, asphalt-saturated felt, and
asphalt shingles.

214

framing members with wallboard applied to both sides. Windows
were of the steel casement type. The exterior was finished with a
special paint, designed to derive added weather protection from
imbedded grains of stone and similar to stucco in appearance. Har-
man did a substantial amount of engineering in adopting various
existing and new materials to the final house, but the company
operated primarily as an assembly plant, purchasing most of its
components elsewhere, and it was notable for the completeness of
the package furnished from the plant. The system inherently re
quired a good deal of relatively skilled site labor and led to unusually
high erection costs.

A metal frame assembly house even better known not so long ago
was the hemispherical Fuller house (see Figure 28). The entire
weight of this structure was borne by a central mast composed of
seven high-strength alloy steel tubes bound together. The mast
rested in a concrete footing, and three steel rings were hung from
the top of the mast, one below the other in widening circles. These
rings changed the direction and fixed the position of the tension
wires which supported the structure. The tension wires were fas
tened to the top of the mast and supported the outer edge of the
circular floor structure, which was composed of wedge-shaped pressed
aluminum floor beams with their narrow ends supported by the central
mast. Curved ribs, acting in principle like those of an umbrella,
supported the roof skin of aluminum. The side walls were curved
double aluminum sheets with space between for insulation. Since
the interior was entirely free of structural members with the excep
tion of the central mast, the room arrangement was quite flexible.
The final house was a metal stressed skin structure, but it was con
templated that it would be shipped knocked down, with its various
framing members and skins packaged into a cylindrical container
4%' in diameter and 16' long, so it is therefore classified here as a
metal frame assembly.

Chart E presents a general summary of the construction details
of the twelve companies using frame assembly in their walls. The
great variety of systems falling under this heading is immediately
apparent.

215

Chan E

Frame Assembly Wall
12 Companies

Frame members

Insulation

Vapor barrier

Exterior structural cladding

Exterior finish

wood

steel

2" X 4"

16" o.c.

24" o.c.

4' o.c.

nailed

bolted

glued and nailed

metal foil

gypslim board

metal foil

paper

asphalt membrane

wood

fiberboard

plywood

nailed

screwed or bolted

glued and nailed

troweled

wood siding

wood cladding

metal cladding

fiberboard

nailed

special

screwed or bolted

troweled

216

Interior surface

gypsum board

lath and plaster

plywood

nailed

troweled

screwed or bolted

glued and nailed

3. Frame Panels

The most common form of frame panel walls consisted simply of
framing members assembled together with exterior surfacing ele
ments. Interior surfacing was commonly field applied in order to
facilitate field installation of plumbing, heating, and wiring and to
simplify the application of acceptable interior finish.

It is interesting to note that two of the largest producers of houses
in this category 26 ( both of which combined pref abrication and site-
fabrication techniques) used stucco for at least part of the exterior
surface. This serves to illustrate the fact that strict rules cannot
be laid down for the prefabrication process, for combinations of fa
vorable climatic conditions, projects large enough to permit efficient
techniques of application (in this case gunite), and highly organized
construction systems can turn to competitive advantage even the wet
processes often considered incompatible with prefabrication. It is
foolish, therefore, to attempt to discover an absolute scale of values
or an ideal to which various systems may be compared. It cannot
too often be emphasized that design must be considered in a very
broad sense, embracing the whole production and marketing scheme
of the prefabricator. Nevertheless, the smallest details of design
have a considerable interest of their own and are also properly the
subject of consideration.

Panel size. Ordinarily the factors controlling size are the bulk and
weight which can easily be manhandled in the field. With the
exception of the few producers of modular panels, whose aim in
brief was to produce a rather more complex variety of stock build
ing material, there has been a noticeable tendency, not clearly re
flected in the data, to use the largest panel practicable in the field,
thereby cutting down on erection labor and field jointing, among

26 Kaiser Community Homes and the Byrne Organization.

277

other things.27 With this in mind, several producers had changed
their 3' or 4' modular panels to room-size, or even wall-size, panels.

Field joints have become to most prefabricators a design problem
of the greatest importance. Many of them firmly believe that verti
cal battens or any other external indication of joints, particularly
on the exterior of the house, will be considered objectionable by a
substantial part of their potential market, and some of them feel that
it is important to conceal from the public the fact that the house is
prefabricated at all. And it must be conceded that there has been
some justification for this concern about the public reaction to visible
joints.

This becomes, then, a limiting factor on the extent of factory
application of surfacing materials, and many prefabricators have
made a practice of leaving finish siding, facing, or shingling as a
field chore in order to conceal joints. The use of room-size panels,
on the other hand, offers the opportunity to conceal, eliminate, or
finish with precision methods in the plant the joints of inside wall
surfacing materials. It also becomes possible to reduce the exterior
joints to one or two per wall and to disguise these by locating them
at natural breaks in the elevation or by concealing them behind
downspouts or other exterior details.

A few companies carried this a step further and produced wall-
size panels, but since these usually require special handling equip
ment at the site, they tend to be limited to specialized situations.
Nevertheless, as finishing materials become commercially available
in larger sheets and as new lightweight walls are perfected (such as
the plastic-impregnated paper-core sandwich walls), this trend to
wards larger panel size will probably continue.

Joints. A great deal of ingenuity has been exercised to develop
joints 28 that will be at the same time simple to produce, hard to
damage in transit or at the site, easy to erect in the field, and satis
factory in terms of performance in the final house. Refinements in
clude joints which permit panels to be put together either way rather
than only one way (such as left to right), three-way joints for places
where interior partitions join exterior walls, and joints which will
connect standardized ceiling, roof, and floor panels as well as wall

27 When Harnischfeger shifted from modular to room-sized panels, considerably
less framing lumber was required, as well as many fewer kinds of parts.

28 All panel systems and many other systems require field joints. Although
they are first discussed here, their applicability is in no way limited to frame
panels.

218

panels. Of at least some importance in selecting a joint system
has been the possibility of getting patents on it.

By and large, the majority of the prefabricated houses produced
during the period of the survey were of wood construction, and the
detailed problems of joint designs adequate for such systems need
not be considered here since they are adequately discussed else
where.29 The importance generally accorded the joint is so great,
however, that a brief summary of the various types is presented
here. These joints are illustrated in Figure 18.

A single or double lap joint is formed by the butting of contiguous
skin sheets, on one or on both surfaces, over a common framing mem
ber to which one skin sheet is usually bonded in the factory and the
other in the field. Such a joint is referred to in the industry as a
male and female joint. The common member may, of course, be a
filler strip which fits into recessed edges on both panels, and in this
case it approaches a spline in character.

Batten strips can be used to join two panels which are butted
together, and they usually increase the weathertightness of the
joint. The batten is one of the oldest and simplest of joint methods,
but there is a marked tendency to avoid it because of the belief that
the general public will not tolerate such a sign of transitory character
in a permanent house.

The spline is usually a continuous joint, and it is popular because
it permits use of the same simple field device at either edge of the
panel and because it permits flush finishing in the factory of both
frame and surfacing materials— an advantage in transporting and
handling and in certain manufacturing and finishing operations.
This involves somewhat more millwork than the previous joints, and,
since most prefabricators prefer not to nail in the spline directly
through the surface skin because this means a nail head or hole to
conceal, the structural tie achieved is often not so strong as in the
case of other joints. The spline joint is, therefore, generally used to
close vertical gaps and line up panels vertically, and rarely if ever to
make horizontal connections.

The interlocking joint is the most complex type of joint, mostly
used for panels having both surfaces applied and finished in the
factory. It requires a dimensional precision which is not easy to
realize in ordinary framing woods, and may involve extensive mill-
work. Because of differential shrinkage in wood, for instance, Z
joints have frequently given trouble. Frequently special metal parts

29 Manual on Wood Construction for Prefabricated Houses, Chapter 13.

219

Double lap
(also called
male and female)
Single lap has only
one face lapping
adjacent framing
member

Butt and batten

Spline

Interlocking

Figure 18. Commonly Used Panel Joints

are designed to speed up or improve field operation, and nearly all
such joints are patented. Occasionally, these joints seem to have
been designed with more proprietary pride than logic, and some
seem to cause more difficulties than they solve.

Interlocking joints may have extra features, however, as when
the design gives the added elasticity of permitting panels to be
attached horizontally as well as vertically and thus makes possible
the production of stock modular panels as a building material for
assembly according to design of the local architect or builder.

Nearly all exterior joints require some sort of caulking to make
them weathertight, and the ideal caulking material is still to be
found. In the case of most materials the edges of the exterior sur
facing itself must be carefully protected as well. Of course, where
exterior joints are concealed by field-applied (or field-finished) sid
ing or shingling, no such problem arises.

Wood frame panels. Examples of this type of construction were
produced by Kaiser Community Homes (see Figure 29). A com
pletely standardized one-story house "chassis" was produced in the
factory (45% of the work) and individualized to some extent in
the course of the field finishing. This chassis consisted of room-
size panels made up of 2" X 4" studs to which %" plywood was
glued and stapled to serve as interior wallboard.30 Windows and
doors were fitted or hung in their frames in the walls in the plant.
The panels were spiked together in the field, and chicken wire was
applied over building paper as a base for the application of an
exterior finish of stucco, to which areas of siding or shingling were
added for variation in appearance. Interior partitions were in some
cases factory-built storagewalls (entirely utilized for closets, shelves,
drawers, and the like) and in some cases stressed skin panels (de
scribed in the next section). Inside surfaces were finished with a
fabric-base wallpaper. At the time of the survey, about 3,000 of
these houses had been built in the Los Angeles area.

Another example was the house built by the Defoe Shipbuilding
Co., the walls of which were frame panels from 4' to 12' in width,
made up of standard 2" X 4" studs 16" o.c., %" fiberboard sheathing,
and plasterboard interior surface, taped and filled at the site to
present a smooth and unbroken appearance. With both surfaces
factory applied, a special joint was necessary, and an interlocking

30 Both gluing and stapling or nailing are frequently used in this way to give
added rigidity and thereby eliminate much of the need of bracing in the walls
and bridging in the floors. The staples or nails serve principally to apply pressure
until the glue sets.

221

joint was used, requiring some millwork. At the site joining was
accomplished by diagonal nailing from the outside, with the final
joints concealed by exterior siding. Aside from the joints and the
factory application of surfacing, insulation, windows, and doors,
there was little to distinguish this house from the conventional wood
frame house. In that sense, it was typical of a large group of pre
fabricated houses, more of which have probably been built and sold
since the end of the war than any other type. Many of these com
panies, while maintaining profitable operations with this sort of
house, are at the same time working on more unconventional designs
for eventual production.

Metal frame panels. Houses using this type of construction were
far less conventional than those we have just discussed (see Figure
19). The widely advertised Lustron house was a good case in point.
Here the frame consisted of steel studs, rolled in special "hat" sec
tions and welded 2' o.c. on both sides of horizontal "hat" section
members, with channel members welded in between the studs as
bracing. Window and door framing were also welded into the
panel, and the whole framing system then got a bonderizing coat
and a sluiced-on protective enamel coat. The interior and exterior
surfaces were steel pans finished with vitreous enamel, as were the
roof and ceiling surfaces, and all were attached in the field except
ing those in the special bay window section, which was factory
assembled. The vitreous enamel finish was available in a variety
of colors, and it offered a relatively permanent, easily cleaned surface.
For heating, the house had a plenum chamber over the ceiling pans,
converting these into a radiant ceiling, and insulation was factory
applied to the inside of the exterior steel pans and field applied
over the top of the plenum chamber. Through-metal contact and
resulting condensation were minimized by continuing the insula
tion between the metal studs and exterior pans. Weather bond
between pans was achieved by extruded gaskets of Koroseal specially
designed to seal the joint between the flanges of the pans. This
house was notable for the completeness of the package offered
through the chain of Lustron dealers, and for the utility room offer
ing bulk storage within the house.31 The standard two-bedroom
house was about 1,000 sq. ft. in floor area, or substantially larger
than the average prefabricated house. The three-bedroom house
contained more than 1,200 sq. ft.

31 The house was basementless and had a floor of asphalt tile over concrete
grade beam and slab.

222

Another metal frame panel house was that built in a 1,200-house
project at Harundale, Md., by the Byrne Organization in a combina
tion of pref abrication and site-fabrication techniques 32 ( see Figure
28). Here Macomber-type steel studs (rolled to give a "hat-shaped"
section) were used, with two opposed sections spaced by welded
tie rods to make a complete wall stud. The complete studs were in
turn assembled into steel frame panels in a shop at the site, a 1"
glass fiberboard laid over the exterior of these panels, and a paper
backed wire mesh pinned through the fiberboard to the panels by a
special welding machine for stucco finish. The paper acted as a
weather barrier, and the insulation near the outside surface pro
tected the steel members from falling below dew point and causing
condensation, while two coats of special aluminum paint on plaster
served as a vapor barrier. The structure was placed on a radiant-
heated floor slab and stucco applied as an outside surface in the
field. Initially, before the company turned to the use of aluminum
paint, vapor-barrier paper with wire lath had been nailed on the
inside to the frame in the shop and plaster applied in the field. Walls
were vented into the roof space to assist in carrying off any vapor
accumulation which might occur. Interior partitions were site
assembled, and the wall panels and roof trusses were welded together
at the site to form a continuous rigid frame structure, far stronger
than required by codes.

A very different type of metal construction was offered by General
Homes. Although this was primarily a frame panel construction,
reliance was also placed in part upon stressed skin principles. The
core of the panel consisted of 0.032" aluminum sheet, shaped into
continuous trapezoidal sections 4" high and 6" center to center.
The surface, inner and outer, was 0.032" aluminum skin bonded to
%" fiberboard sheets. This surface was screwed to the shaped core
in sheets 2' o.c. through aluminum trim strips. Aluminum straps,
2' o.c. and riveted to the core horizontally, added stiffening. Panels
were locked together in the field by tabs punched out of channels in
the end of the panels. In the shop 4" batts of insulation were in
serted into the cores, and all metal surfaces shop coated with zinc
chromate primer. A simple screw attachment in the field was used
to install windows in special framing prepared in the shop.

32 The recent financial troubles of this project have caused a great deal of
speculation about the principles of location, site planning, design, and fabrication
used by the Byrne Organization. Needless to say, blame cannot be firmly fixed.

223

Porcelain

enamel

gutter

Porcelain
enamel
ventilated
eave panel

Porcelain
enamel
roof panel

Top stud
spacer

Roof truss

Vertical wall
studs

Koroseai
gasket

2-0"x2'-0"
porcelain
enamel
exterior panels

Anchor bolt

Bottom stud
spacer

Premolded
asphalt
joint strip

8" concrete
foundation wall

Porcelain
enamel
gable end
panel

Insulation
supports

Floor slab

Figure 19. Metal Construction Systems: ( 1 ) Lustron

Asbestos shingles
Building paper
Steel gutter

Hat shaped
section
roof truss

Hat shaped
section
steel studs

Steel base

Steel stud
spacer

Asphalt tile
floor

Rigid
insulation

Plywood
sheathing

Fiberglas
insulation

\ Paper
backed
wire mesh

^Stucco
Insulation

Lower truss
member

Metal lath
Plaster
Vapor seal

Radiant heating coils
'in concrete slabs

Figure 19. Metal Construction Systems: (2) Byrne Organization

Chart F gives a summary of the construction details of the 62
companies using frame panels in their walls. It will be noted at
once that the great majority, 55 out of 62, used wood as their princi
pal material, and that, of these, the great majority, 46 out of 55,
used conventional 2" X 4" studs, 16" o.c.

Chart F

Frame Panel Wall
62 Companies

Frame members

Panel size

Joint

Insulation

Vapor barrier

wood

steel

aluminum

2" X4"

2"X3"

16" o.c.

24" o.c.

nailed

welded

glued and nailed

48"

room size

wall size

butt

interlock

batten

fiberboard

batts

metal foil

insulation backing

paper

226

Exterior structural cladding

Exterior finish

Interior surface

wood

plywood

fiber board

steel

aluminum

1* (wood)

M* (plywood)

H" (plywood)

nailed

glued and nailed

shop

field

wood siding

wood cladding surface

metal cladding surface

shop

field

gypsum board

optional

plywood

fiberboard

nailed

troweled

glued and nailed

field

shop

4. Stressed Skin Panels

Lumber and plywood panels. Companies using this system of con
struction varied the most, as a group, from conventional structural
design, and they represent one of the most significant developments
in the field. In all, 32 companies used plywood stressed skin panels.33

83 It has been noted above that the term "stressed skin" is somewhat loosely
applied to this construction, since there is some structural reliance on the framing
members themselves. The Forest Products Laboratory prefers the term "stressed-
cover panels" in this connection, and other terms are undoubtedly in use. "Stressed
skin" is nevertheless retained for use here because the term is so widely familiar.

227

The general characteristics of this system include smaller con
sumption of material and lighter weight, compared to conventional
frame systems, achieved at the same time as increased strength and
structural stiffness. Basically, the system depends on the strength
developed by bonding rigidly together by means of specialized
gluing techniques a system of light wood framing members and thin
plywood surface sheets, so that the whole acts together in the nature
of a box girder. Such a panel, when correctly made, is much
stronger and stiff er than conventional wood frame construction.
Wall thickness can be substantially less than that of conventional
construction. From the viewpoint of economy, the system not only
uses less wood than conventional construction, but has the further
advantage that plywood uses a larger percentage of the log than can
be made into lumber. On the other hand the quality of the framing
wood must in most cases be better. Research has developed water
proof, high-quality glues and improved techniques for binding the
surface sheets to the framing members, and current development
work on the composition and design of these sheets indicates that
further improvements are to be expected in the future.

At the present time, the quality of the construction is such that
the plywood surfaces can be used for interior and even exterior
finish, although there is still some difficulty in maintaining good
exterior finish on plywood exposed to the weather. In this respect,
edge-grained plywood and the recent redwood plywood perform
better, and plastic and other surface coatings further improve per
formance. The framing members, because of their smaller section
and the need to present a flat surface for gluing, are somewhat higher
in quality than ordinary framing and in some cases can be used
by prefabricators as actual floor and window framing, thus vastly
reducing requirements for trim in these locations.

Interior partitions are frequently of the same construction, even
when not load bearing, because of the availability of manufacturing
facilities, although single plywood sheets, practically self-support
ing, are occasionally used in locations such as closets where the
sheet is not the sole barrier between rooms.

The manner in which some of the new technical problems pre
sented by stressed skin plywood panels have been met by the pre
fabricators is of interest, particularly since in many cases these prob
lems are also faced in the use of other new materials. For that rea
son, a few of the most significant problems will be discussed here
briefly.

228

Condensation. A construction which makes the wall markedly more
airtight creates the benefit of diminished heat loss but also the prob
lem of diminished vapor permeation, and therefore condensation.
When the vapor originating within a house in cold weather cannot
readily escape through the floors, walls, and roof (as is the case
with stressed skin plywood or with metal panel walls, for example),
there is always the possibility of its coming in contact with some
frame or surfacing element which has been cooled by outside air to
a temperature below the dew point. At the point of such contact
condensation occurs, and as this point is likely to be within the wall
structure itself, the resulting moisture may do considerable damage.34
Among the prefabricators using stressed skin plywood and metal
panels, at least nine were using special ventilation slots within the
wall space, and many used vapor barriers designed to reduce the
penetration of vapor into the wall structure from the house area.

Among the most commonly used vapor barriers were asphalt mem
branes laminated with kraft paper (and frequently backed by in
sulation) and metal foils. The performance of such barriers is good
only when they are carefully fitted and tightly fastened in place.
In the common examples in which the material supposedly serving
as a vapor barrier was not even tacked or stapled in place, there was
serious question of its effectiveness.

A problem of detail in stressed skin construction is that the heads
of nails or staples on the interior surface are frequently the points
at which there may be condensation, resulting in dark spots or stains.
Several companies sought to solve this problem by countersinking
and puttying over these heads, while others avoided it by bonding
the plywood to the frame without nails or staples, through the use
of glues set with hot or cold presses.

In stressed skin plywood, as in plaster and many other materials,
a combination of such factors as the static charges on dust particles
and the differential rate of thermal and vapor conductivity between
the sheet alone and the sheet backed by framing will often result
in the collection of dust and dirt in such a way as to show on the
surface the pattern of framing in the form of so-called "shadow lines."
Some prefabricators sought to avoid this effect by developing designs
which tended to equalize thermal and moisture conductivity through-

34 For a discussion of the condensation problems in some fifty different wood
and metal wall and roof constructions tested in the Pennsylvania State College
Climatometer see Ralph R. Britton, "Condensation in Walls and Roofs," HHFA
Technical Paper, nos. 1, 2, 3, and 8 (March, June, and September 1947; April
1948).

229

out. Gunnison, for example, fluted the framing members wherever
they came in contact with the surface skin in an effort to reduce this
differential.

Insulation. The only common characteristic observed among com
panies in this respect was the almost complete absence of loose or
fill-type insulation for walls. The common insulations were in the
form of sheets, batts, or blankets, usually paper backed; and the
considerations to be met in the selection of an insulating material
included the ease of handling and installation, the durability, the
insulative quality, and the amount of space available for storage
in the plant.

Finishes. As a result of the development of waterproof glues,
delamination of plies is a rare occurrence if edges are well protected,
and of the 32 companies using stressed skin plywood construction
in their walls, 20 used the plywood as the exterior finish material.
A good deal of development work in paints and sealers has helped
to make this possible,35 although some of the good sealers, such as
aluminum flake, have been little used by prefabricators because they
add production difficulties.

Several of the large companies were using the relatively new de
vice of bonding plastic-impregnated paper to the plywood to serve
as an exterior surface material and an excellent base for paint. Other
companies made use of striated plywood, the scoring of which tends
to conceal any checking which may occur, to disguise the joints
between panels, and to give a pleasantly textured finish. On inte
riors, where neither of these devices was used, it was common to
find a rounding or beveling of the edge of the plywood sheets, in
recognition of the fact that expansion and contraction of the panel
surfaces will otherwise eventually make visible cracks in paint or
paper surfacing anyway.

Nearly all the companies using stressed skin plywood panels
without further interior or exterior finishing materials designed their
houses in terms of modular panel widths of 40" or 4', principally
because the plywood sheets come in widths of 4', but partly also
because of the tendency of larger panels to bulge as the exterior and
interior sheets develop large differences in moisture content. Where
additional surfacing materials are used, this tendency can be rather
easily controlled.

A few of the companies prefinished the plywood interior in the
shop, usually with tinted sealers and lacquers to produce a subdued

85 See Manual on Wood Construction for Prefabricated Houses, Chapter 7.

230

grain finish. Most companies finished in the field, taking special
precautions to combat the cracking of the finish materials at the
joints. One method used for this purpose was taping and puttying,
and fabric-based or other special wallpaper was another. Some com
panies used batten strips, taking precautions to avoid the opening of
cracks in the finish along the batten edges which result from shrink
age and from movement of the panels behind them.

Large plywood sheets were not ordinarily used in stressed skin
panels, but more usually in frame panel construction. Such sheets
were made up by factory joining of 4' X 8' sheets into room-size
sheets by scarf or lap joining and gluing under a hot press, yielding
a very satisfactory continuous wall.

This brief discussion of the technical problems of plywood will
illustrate the fact that the use of the material for efficient home
construction has depended in large measure upon the development
of sound factory processes. In the opinion of many a prefabricator,
it is definitely a prefabrication material, which could be used effec
tively in the field only through craftsmanship of cabinetmaker
quality.

Several of the companies used laminated paper Upson board as
an interior wall surface material, and at least one used it in a partially
stressed skin design. Most frequently this material was used in
room-size sheets, with the openings cut out of it. Designs took care
to conceal cracks or unpainted lines which might appear as the
result of movement of the board over the course of time. Other
materials were also employed, such as Homasote, a wood-pulp board
available in room-size sheets, used in one case in partially stressed
skin construction.

Examples. Green's Ready-Built Homes, Ivon R. Ford, Inc., and
Winner Manufacturing Company, Inc., all produced stressed skin
panel houses. Green's Ready-Built Homes produced a panelized "solar
house" (see Figure 30), of which much of the design was the work
of George Fred Keck of Chicago. The basic panel was 39" wide,
of wall height, and composed of %" exterior grade and %" interior
finish plywood glued (by high-frequency induction hot press) to a
frame of 2" X 3" edge members supported by two 1" X 3" inter
mediate studs. Panels were held together by metal connectors on
the grooved and beveled panel edges, the connectors being held
firmly by the position of the heads of screws attached to the inter
secting edge of a partition or corner panel (and also to battens,
which were inserted at each joint). This gave the system a feature
of demountability. Between the plywood surfaces two aluminum-

231

foil reflective layers were carefully bonded in such a way as to
create three separate and approximately equal insulating air spaces
within the panel. Exterior surfaces were sealed, primed, and fin
ished with oil paints, and interior surfaces were shop finished with
clear stain and lacquer or paint, as required. At the site wall panels
were positioned in an extruded aluminum plate. A feature of this
house, and one which was beginning to appear elsewhere also, was
the use of fixed and sealed double panes of glass for vision, with
separate wood panel openings top and bottom, louvered and screened,
for ventilation. This design permitted the use of larger windows with
out the expense of the complex carpentry often encountered in mov
able sash construction. The house had a high degree of factory
finish, and the finish was of a quality which would be expensive to
duplicate in the field. In architectural planning the house was also
unconventional, with an attempt to give all rooms due south orien
tation.

I von R. Ford, Inc., and its nine licensees spread over the United
States and Canada manufactured stressed skin plywood houses (see
Figure 35) made up from room-size panels of %" exterior sheathing
and %" interior finish plywood glued and nailed to 2" X 3" studding
16" o.c. Doors and double-hung windows were factory installed.
Joints, of male and female type, were glued as well as nailed, and
siding or shingling was applied over the plywood sheathing, in the
field. The bottom plate of the wall panel was rabbeted into the sill
for alignment and fastened in place by toe-nailing into the sill with
spikes. The design was relatively conventional and simple, and yet
would be difficult to produce in economical quantity without a well-
equipped woodworking shop.

The Winner Manufacturing Company at the time of our survey
produced the Shelter Industries House, designed by Donald Desky.
Room -size panels were made up of %" striated exterior plywood
and %6" striated interior plywood bonded to IKe" X 3%" studs
16" o.c. Joints required to make up plywood sheets of the required
size were factory caulked. Double aluminum-foil sheets and one
vapor barrier were suspended in the wall to give a total of four
interior air spaces. Field joints were of the male and female type,
nailed and caulked. Doors and double-hung windows were factory
installed, and the striated plywood finish was field finished with
lacquer or paint as desired. This house was of unconventional archi
tectural as well as structural design.

Metal skin panels. This type of construction was used by the
Butler Manufacturing Company which made a house (see Figure

232

36) of 2' wide aluminum pans of wall height which acted in the
manner of vertical channels, flange to flange. The web of the chan
nel served as the exterior surface of the house, and in structural action
constituted a stressed skin 0.051" thick. The 4" flanges gripped wood
filler strips which cut down the thermal conductivity from outside
to inside. In part, also, these filler strips served as a frame, and they
provided a wood surface for the nailing of optional interior finish ma
terials. Two braces per panel stiffened the flanges, and the panels
were clipped together at the site with an H-shaped key, driven home
by a hammer. Blanket insulation was added to the reflective insula
tion provided by the aluminum surfaces themselves, and panels were
ventilated to the outside to minimize condensation and obviate the
necessity of a vapor barrier. Window sash and frames were of ex
truded aluminum sections, and doors and door frames were of wood.
Shorter lengths of panel were used under and over openings and to
make up gable ends. Aluminum channels positioned the panels top
and bottom and were bolted into the floor and ceiling; wood molding
and base were used on the interior. Exterior finish was paint applied
as desired in the field over shop-applied zinc chromate primer.

Plastic-impregnated paper-core materials (see Figure 20). One of
the most interesting technical developments in the postwar period,
this construction was originally a development of the aircraft industry.
Used because of their high strength/weight ratio, such materials have
been the subject of a great deal of interest and investigation on the
part of the prefabricators. Usually classified as sandwich construc
tion, they are generally most like the stressed skin construction of
those classifications which are used here. The two forms most com
monly seen at the time of our survey were the honeycomb paper core
and the corrugated paper core.36 Development work on the former
was carried on primarily by Lincoln Houses Corporation, Chrysler,
Douglas Aircraft, and Consolidated Vultee; while the latter has been
developed primarily by the U. S. Forest Products Laboratory.

At the time of the survey, Utley-Lincoln planned to make a house
with panels of the Lincoln core and aluminum skins bonded by the
Chrysler method. Southern California Homes was starting produc
tion on a house of similar basic material, and it is an indication of the
structural advantages of this development that, for southern California
climatic conditions, a wall section only 2" thick required no further in
sulation.

86 Described in more detail in Chapter 9.

233

Typical structural panel

Honeycomb paper core

STEP 1.
Glue applied
at alternate
lines on paper

1. Cross-corrugated
parallel to faces

2. Cross-corrugated
perpendicular to
faces

Cross-corrugated paper core

Asbestos
cement

3-ply
Celotex

Asbestos
cement

STEP 2.
Paper pressed
together and
allowed to cure

STEPS.

Then pulled apart
to form irregular
pattern of cells

Cemesto board
(non-structural paneD

Figure 20. Sandwich Panel Materials

The Southern California Homes system included a semihoneycomb
paper core, impregnated with phenolic resin, and faced with alumi
num skins (3S, 0.020" thick, % hard) in room-size panels. Panel
edges and openings were formed by channels of aluminum 0.064"
thick, the flanges of which were bonded between skin and core ma
terials. The bottom edge of the panel was bonded to a 2" box sec
tion of aluminum which served as a wiring conduit and also pro
vided access every 4' for bolting the panels to the foundation slab.
On the inside this was covered by a simple baseboard. Aluminum
rolled strip door and window stops were screwed to the channels,
and served to locate steel casement sash and paper-core wood-veneer
flush doors. On the outside the bottom edge of the panel had a
lip to cover the exterior joint at the edge of the foundation. From the
top edge of wall panels, bolts passed through to roof panels and
held them in place. Paint over zinc chromate primer was the finish.
This wall, complete, averaged only 1 Ib. per sq. ft. Architecturally,
the house was of unconventional design and reflected an appreciation
of the possibilities of the new material. It would require a few
changes, however, particularly to avoid through-metal in the walls,
for use in northern climates.

Steel skins, particularly stainless steel, could of course be used as
well as aluminum in paper-cored walls, but as yet no prefabricator
had tried them. There was during the time of the survey, however,
at least one house making use of plywood-faced paper-core walls.
This was the Acorn House, designed by Carl Koch and John Bemis.
Here the design called for a core, not of the honeycomb type, but
made up of corrugated paper with the direction of the corrugations
alternated for good performance structurally and as insulation, ac
cording to the recommendations of the U. S. Forest Products Labora
tory. The core was then bonded in a press to surfacing sheets of %"
interior grade plywood and %" striated exterior plywood. Ad
vantage was taken of the lightness of the material to design a house
of 800 sq. ft. which could be put together in the factory around a
completely equipped kitchen, bathroom, and utility core, folded into
a compact 9' X 24' unit, shipped to the site on an ordinary trailer, and
there set on posts and unfolded in a very simple operation to produce
the finished house. Folding was made possible by simple hinged
joints sealed with neoprene gaskets. This house, light as it was, was
designed for use in northern climates.

The development of these new lightweight materials has made it
possible to assemble larger and larger sections, thus avoiding the field
joint and site labor problems. This tendency in turn has brought

with it trends towards a greater degree of standardization in the final
house, in order to make possible simple and repetitive factory oper
ations, and towards a highly integrated design for the whole house
which assures full benefit of the thermal, structural, and acoustical
properties of the new materials and careful attention to technical pos
sibilities and difficulties.

Chart G

Stressed Skin Panel Wall
41 Companies

Frame members
or structural core

Panel size

Joint

Insulation

wood

aluminum

steel

plastic-impregnated paper

2" X3"

1" X4"

2"X2"

16" o.c.

VA" o.c.

12" o.c.

glued and nailed

nailed

glued

48"

room size

40"

m and f

spline

butt

batts

blanket

metal foil

236

Vapor barrier

Exterior structural cladding

Exterior finish

Interior surface

insulation backing

spray or brush coat

metal foil

plywood

wood

fiberboard

aluminum

steel

3^"

glued and nailed

electronic glue

hot press glue

nailed

shop

field

plywood cladding surface

wood siding

metal cladding surface

shop

field

plywood

gypsum board

aluminum

glued and nailed

electronic glue

hot press glue

nailed

shop

field

Chart G gives a summary of the structural details used by the 41
companies which were, at the time of the survey, making use in
their walls of stressed skin panels in some one of the many forms
which have been considered.

237

5. Solid Panels

Precast concrete. The solid panel wall, as the term has been used
here, is a wall in which the structural loading is spread throughout
a more or less homogeneous panel system. The most common design
falling within this classification is the "precast concrete" wall, of which
there were 12 types under development at the time of the survey.
At the time, very few were in production, and none on a large scale,
although wide interest was evinced in their possibilities.

Some of the disadvantages of concrete for this purpose have al
ready been mentioned on p. 183. Of these, perhaps the greatest is
the combination of weight and bulk with frangibility at the edges,
although the designer must also deal with poor thermal insulation and
with the problem of obtaining an adequate degree of accuracy for
any complex concrete shape at the same time as production and erec
tion economy. Lightweight aggregates and foamed concretes offer
a potential saving in weight and an improvement in thermal insula
tion, but relatively few prefabricators were making use of them. In
deed, most concrete systems were designed for relatively simple and
unfinished houses in relatively warm climates.

The hollow slab is, of course, a partial answer to the problems of
heavy weight and poor thermal insulation, but it is also much more
difficult to precast than a solid slab. For low cost, the Portland Ce
ment Association suggested precast ribbed slabs, and a few of the
prefabricators used them. In most precast concrete systems, care was
required to provide thermal insulation and to avoid condensation, par
ticularly along the joints where there was likely to be through-con
crete. At the joints, also, provision had to be made for expansion and
contraction in the heavy concrete masses, and for the problems of
displacement and poor alignment resulting from inaccuracies or
changes in dimension or shape along the slab edges. An elastic
joining material of relatively great thickness usually was employed
for these reasons. It was also necessary, in most cases, because of
the weight and bulk of the panels, to use additional reinforcing ma
terial to protect them during the various handling operations from
casting to final placing.

Nevertheless, precast concrete slabs offered certain definite ad
vantages, such as overall design simplicity and resistance to com
bustion, corrosion, and insects. Development of simple casting pro
cesses, of concrete aggregates permitting easier handling and simpler

238

design, and of more effective joint details may well bring further con
centration on concrete in the next few years.37

An example of a precast concrete house was that being developed
by Merriam and Twachtman at the time of the survey. Wall panels
of room size, as large as 8' X 20', were to be precast of a vibrated
concrete incorporating an expanded slag aggregate, and given an
exterior surface of white cement which was supposed to require no
further painting. The interior surface was to be a factory-applied
plastic-base paint intended to serve as a vapor barrier as well. Re
inforcement was to be provided by wire mesh and by bars tightened
together at vertical joints to tie the panels together and hold them
in proper alignment. Thick strips of rubber mastic were to be used
at all joints, allowing for imperfections and cutting down acoustic
and thermal transmission; door bucks composed of metal sills and
jambs were cast in at the time of making the sections, and exterior
wall joints were to be protected by precast pilasters.

Other precast concrete systems were quite different in concept;
for example, there was the "Pfeifer Unit" produced by The U. S.
Housing Materials Corporation. This was small (24" X 24") and
served the function more of a building block than of a true wall
panel. The most familiar example of the building block, of course, is
the standard concrete block, made by simple machinery and available
in every part of the country. Probably neither block should really
be classed as a wall panel.

Composite materials. These may also come under the heading of
solid panel walls. Cemesto Houses have already been described
under frame and curtain wall assemblies, but the use of a similar
board as a load-bearing panel without structural framing has also
aroused some interest, although it would require a tougher skin ma
terial and careful attention to the joint and to the waterproofing of

37 The work of Corwin Willson on concretes of extremely varying characteristics
was reported during World War II. Willson tested 8,000 specimens from 225
materials, which included organic and mineral wastes of all sorts in combination
with a variety of fillers, leaveners, stabilizers, waterproofers, and surface coatings.
He found potentially useful combinations composed of common wastes, and he
was able to secure many concretes of excellent characteristics for building con
struction. The possibility is thus presented that even a manufacturer of wood
houses may make extensive use of his waste products for such purposes as floor
tiles, wall coverings, piers, slabs, or other building components not now supplied
in his house package. (Corwin D. Willson, Properties of Assorted Light Weight
Aggregate Materials, Office of Production Research and Development [Washing
ton, 1944]. Also available from Hobart Publishing Company, Washington, D. C.)

239

Chart H

Solid Panel Wall
12 Companies

Frame members
or structural core

Panel size

Joint

Insulation

Vapor barrier
Exterior finish

Interior surface

concrete

steel

aluminum

6" thickness

4" thickness

poured

bolted

glued

room size

48"

butt

tongue and groove

insulating concrete

loose

spray or brush coat

membrane

metal foil

concrete cladding

stucco or concrete

shop

field

cement concrete

lath and plaster

gypsum board

field

shop

240

the panel edge, since the Celotex filler of the Cemesto sandwich loses
as much as 80% of its shear strength when it is wet.

Another familiar composite which may eventually be used in this
way is the panel of wood chips bound with cementitious materials.
This has been experimented with for years, and such a panel was be
ing manufactured for industrial insulation by the Porete Mfg. Co.
under the name of Porex. Durisol, originally manufactured in Switzer
land and widely used in Europe, is a similar product recently made
available in the United States. Material of finer wood filler and
denser composition was proposed by Newark Industries of Ohio.
These compositions, although simple to prepare in small amounts,
present a very difficult mechanical problem of mixing and handling
for continuous strip production, and many of them can be thoroughly
protected against moisture only through the exercise of great care.
They were not in use for prefabrication purposes at the time of our
survey.

Chart H presents construction details of the 12 companies making
use of one or another of these solid panel walls when our survey was
made.

6. Poured at Site

While this construction is not prefabrication, as we ordinarily think
of it, some mention should be made of industrialized systems which,
through design and factory fabrication of standardized forming and
pouring equipment, amount almost to prefabrication. In their most
extreme development, as in the LeTourneau house, such systems
achieve mechanization and standardization comparable to those of
the most standardized factory-built houses.

Hundreds, and perhaps thousands, of systems for more efficient
site pouring of concrete walls have been developed, and from the start
it has been apparent that ease of forming is of paramount importance.
One of the simplest solutions has been the so-called "tilt up" system,
in which the walls are cast horizontally, often directly on the floor
slab and sometimes in tiers one on top of the other, and then tilted
up into vertical position by various ingenious mechanical contrivances.
Use has been made of wood blocks during pouring to position the re
inforcement and in the finished wall to serve as nailing blocks. The
potential savings over vertical casting are obvious. Important work

241

on systems of this sort has been done by the Portland Cement Asso
ciation, by E. J. Rappoli, F. N. Severud, and many others.

Hal B. Hayes, producer of the Hayes Econocrete House, has done
a great deal of work on the West Coast on industrialized collapsible
forms for standardized houses, including withdrawable cores for
cavity wall construction. However, at the time of the survey, he
was more interested in factory precasting, in lightweight waterproof
concrete, of solid wall and roof slabs 2%" thick and room size, with
interlocking corners and tongue and groove joints.

A mixed-material system, studied shortly after World War II by
A. J. Higgins, made use of chemically foamed insulating concrete
poured at the site into permanent "forms" of vitreous enameled steel
which served as surface finish. In later development work, Higgins
abandoned the steel surfaces and was at work on the design of a
house using only site-poured foamed concrete slabs.

One of the most ingenious systems, in terms of structural efficiency,
was that of the Neff Airform house. This was not prefabricated, but
made at the site of gunite (concrete) sprayed on light wire-mesh re
inforcement which had been laid over an inflated rubberized-fabric
form. The resulting structure was a monolithic monocoque which
could be hemispheric, ellipsoidal, or semicylindrical in shape, and its
efficient use of materials and form offered many potential cost sav
ings. Neff produced many houses, mostly abroad, but there were
difficulties in carrying out his system, particularly in construction
controls and in making window and door openings and connections
between shapes of this sort.

The LeTourneau system was one of the most elaborate house-cast
ing schemes to be commercially marketed and developed.38 It in
volved the use of a special forming system and a tremendous ma
chine, the Tournalayer, which could pick up a monolithically cast
house, carry it to the site, and position it there for final finishing
(see Figure 37).

Another site-poured system of considerable interest was that of
the Ibec Housing Corporation (see Figure 38). This system, de
signed to apply mechanical processes to large-scale concrete construc
tion of houses, used heavy lifting equipment and a unitary set of
wall forms. These forms, operated by levers and having nylon-
rubber corners to speed up and facilitate stripping and resetting,
were used on a 24-hour cycle. Roof slabs were poured on the ground
in stacks, pancake fashion, thus requiring only edge forms and per-

88 This system is fully described in Chapter 9.

242

mitting the top surface of each slab to act as the form for the suc
ceeding slab. After the form for the walls and partitions had been
lifted, the roof slab was picked up and set in place by a vacuum
lifting mat. Ibec used lightweight aggregates for wall and roof con
struction where climatic conditions required an insulated wall.

It should be noted in connection with these various systems that, in
the first place, all have their most likely application in large projects
and, in the second place, all require a great deal of finishing work at
the site, often by handicraft methods, to produce a house having the
heating, lighting, and mechanical standards common to this country.
In warm climates particularly, and in special circumstances of urgent
need for shelter or lack of other housing materials, however, these
systems have already proved their usefulness and offer real advan
tages.

7. Windows and Doors

One of the important considerations in the prefabrication of walls
has little to do with the structure of the wall section itself; this is the
problem of windows and doors. In the course of the earlier discussion,
reference has occasionally been made to the manner in which win
dows and doors are incorporated into the walls. Here attention is
turned to the methods of fabrication of these elements.

There was an increasing popularity of metal windows, in wood as
well as metal construction. This was largely because of the superior
dimensional stability of the metals, which helps to give an accurate
and lasting fit. However, at the time of the survey, the metal win
dows, particularly aluminum, were higher in initial cost than the
wood windows and were able to compete principally because of di
mensional stability and the possibilities of lower overall cost of main
tenance. Undoubtedly design factors, consumer preference, and sta
bility of supply also affected the choice. Counteracting these ad
vantages were the problems of condensation on frames and sash,
particularly on frames in the wall interior, and of substantial heat loss
through the metal frames.

Windows have long been prefabricated by specialty companies, as
have doors. At least seven of the prefabricators had developed new
window designs, however, to fit their particular houses and manufac
turing operations; although they rarely produced these windows,
they usually helped to pay certain costs of tooling up, such as the

243

cost of an extrusion die for an aluminum window. Particularly when
the wall section was thin, there was an effort to cut down the section
of the windows by having the glass slide or roll without any enclos
ing sash in a frame carefully designed to avoid stress concentrations
on the glass. This usually required better-quality glass and special
attention to details, however, and therefore did not give lower final
costs.

Many of the manufacturers, particularly those using frame panels,
tended to omit weatherstripping as such from wood window construc
tion, although some used metal sash guides and thereby obtained
some weatherstripping action. The metal windows were usually de
signed to give full weatherstripping action.

In the case of doors, there was a definite trend towards the use
of lightweight composite doors with various types of grids for cores,
including plastic-impregnated paper. Such doors were especially
common among the makers of stressed skin plywood panels, for they
are just another form of stressed skin plywood or sandwich panel.
At least one of the prefabricators in this group regularly used the
plywood cut-outs from the door openings in his wall panels to make
up flush doors with light wooden grid cores on his own presses.

One development in this connection which deserves mention was
the use by the Lustron Corporation of sliding doors in most interior
locations in its house. Such a door, designed to be foolproof, has
obvious advantages in the saving of space, and public reaction to
it in the Lustron houses seems to have been favorable. Made up as
a part of a storagewall system of interior partitions, this may be the
first of a number of such doors to appear in prefabricated houses.

E. Ceilings

1. General

The ceiling is one element of the house which should lend itself
readily to mass production because of its large unbroken surface, and
which should be prefinished more accurately and easily in the shop
than in the field. The elimination of the awkwardness and difficulty
of conventional ceiling construction is a natural goal for the prefabri
cators.

244

Furthermore, the ceiling offers to ingenious designers an oppor
tunity for improvement of house performance with respect to acous
tics, lighting, and heating. Many heating engineers and physiologists
have argued that the best position for radiant heating is in the ceil
ing surface, despite the fact that other considerations have led to
the installation of most radiant heating systems in the floor at the
present time. Modern lighting experts are turning more and more
attention to the overall luminous ceiling; acoustics experts have long
recognized the overwhelming importance of the ceiling for control
of sound. While in some respects these goals may be opposed to
one another, they offer a special opportunity and a challenge to the
prefabricator.

Structurally, one of the most important considerations regarding the
ceiling design is the usual code regulation limiting maximum deflec
tion. Few prefabricators use plaster, which is ordinarily allowed a
maximum deflection of %60 of the span, while many use dry finishes
which are allowed maximum deflections up to %4o-89 Further, the
dry finishes reduce dead load and thus permit further reduction in
the framing members supporting the ceiling. This saving is well
enough understood by the prefabricators, and they have fought to
have such construction allowed in building codes.

Other considerations, of course, are the possibility of hanging an
extremely light ceiling from the roofing frame; the possibility of using
as the ceiling merely the underside of the roof construction system,
as would be the case in many flat-roof schemes; and the possibility
of designing the roof structure to be supported only on exterior walls,
leaving the ceiling surface unbroken by load-bearing partitions.

For the most part, it will be seen that the prefabricators were con
servative in their thinking on these matters, with the exception of the
last. The bulk of the designs were quite conventional, and, as a re
sult, the ceilings were often the least prefabricated components in
the house.

2. Frame Assembly

At the time of the survey, 44 companies used frame assembly sys
tems in constructing their ceilings. In 23 cases, the frames which
supported the ceiling were also the bottom chords of the roof trusses;

39 Prefabricated Homes, Commercial Standard CS 125-47, p. 11.

245

Chart I

Frame Assembly Ceiling
44 Companies

Frame members

Surface element

Insulation

Vapor barrier
246

wood

steel

aluminum

2*X6"

2"X4/r

2*X8*

16' o.c.

24" o.c.

nailed

screwed or bolted

field

shop ,

part of roof construction? no

yes

bottom chord of truss? yes

gypsum board

plywood

lath and plaster

optional

fiberboard

nailed

troweled

screwed or bolted

field

shop

batts

loose

metal foil

backing of insulation

asphalt membrane

this left for field application only the surfacing sheets themselves, and,
because in most cases the roof trusses were preassembled, field appli
cation of surface components was probably the only practical answer.
Room-size or house-size ceiling sheets would be too difficult to
handle under these conditions, and among the most satisfactory ma
terials used was regular gypsum board.

A good example of unconventional ceiling construction of the frame
assembly type was that of the Lustron house, in which vitreous
enameled steel pans were screwed to the bottom of the doubled
lower chord of the roof truss. Above this, and fastened to the upper
member of this doubled lower chord, were sheets of insulating fiber-
board, topped by heavy insulation. The area between the two sur
faces thus became a plenum chamber which was used to heat the
house, with the steel ceiling pans serving as a radiant heating panel.
Heat was directed throughout the plenum by a sheet-metal baffle sys
tem. This was a most ingenious way of taking advantage of the
characteristics of the Lustron construction system to produce up-to-
date heating; from the point of view of the prefabricator the troubles
with it were the amount of site labor required to put together with
sufficient tightness the many elements of this plenum chamber in the
difficult working area formed by the chords of the roof trusses, and
the risk of high heat loss if this work were not well done.

The details of the construction of frame assemblies used by the
prefabricators in their ceilings are summarized in Chart I.

3. Frame Panels

The use of frame panels in ceilings is complicated by the difficul
ties of installing vapor barriers and insulation and of getting tight,
good-looking joints which are proof against the soiling action likely
to result from the passage of air from the attic space. Nevertheless
54 of the companies in our survey used such ceiling panels in one
way or another, and the summary of the details of their construction
systems is presented in Chart J.

The difficulties with joints have led in general to simplification of
ceiling panel joints and the use, in many cases, of simple lap joints be
tween ceiling surface sheets over a single solid framing member.
Where a regular butt joint was attempted, there was almost no way
to allow for such shrinkage of surface sheets as might have taken

247

place even before installation of panels— other than by the use of
battens, and these were generally felt to be unpopular with the pur
chasing public. Several of the companies had attempted to panelize
such frangible surface materials as gypsum board, but many more
of them used plywood for this purpose.

Chart J

Frame Panel Ceiling
54 Companies

Frame members

Panel size

Joint

248

wood

steel

concrete

aluminum

2" X 4"

V X 6"

2"X 8"

16" o.c.

24" o.c.

8M" o.c.

nailed

glued and nailed

welded

shop

field

part of roof construction? no

yes

bottom chord of truss? no

yes

4' X 12'

8' X 12'

4' X house width

room size

butt

m and f

interlock

Surface element

Insulation

plywood

gypsum board

fiber board

lath and plaster

nailed

glued and nailed

troweled

electronic glue

shop

field

batts

loose

metal foil

fiberboard

insulation backing

metal foil

asphalt membrane

spray or brush coat

Vapor barrier

Experimental panels of this sort were designed by the Crawford
Corporation, using %" plywood in 4' X 12' sheets made up in a
panel with the ceiling joists, and having an impermeable reflective
insulation installed in the plant. A design feature was the use of
steel straps to tie these panels together over the center bearing par-
ition, to tie the panels to the rafters, and to position the structural
elements before nailing.

4. Stressed Skin Panels

Because of the excess strength inherent in the stressed skin panel,
it is generally used for the ceiling only when it also serves as the roof
or as a second floor designed to carry live loads. In these circum
stances it offers some advantages in the reduction of dead loads, and
in fact it may be more appropriate than on the first floor, where there
is rarely any advantage in a smooth undersurface.

However, it is difficult to avoid irregular-appearing joints between
such panels without using special batten or jointing strips, unless the
complete ceiling for each room area is made up as a single panel.

249

Chart K

Stressed Skin Panel Ceiling
21 Companies

Frame members
or structural core

Panel size

Joint

250

wood

aluminum

plastic-impregnated paper

steel

2" X 6"

2*X4*

2"X3"

16" o.c.

12" o.c.

8}i" o.c.

glued and nailed

hot press glue

nailed

screwed or bolted

electronic glue

shop

field

part of roof construction? yes

bottom chord of truss? no

yes

4' X 12'

4' X house width

40" X 10' or 6'8"

4'X4'

4'X8'

8' X house width

butt

"v"

spline

m and f

interlock

Surface element

plywood

aluminum

gypsum board

steel

glued and nailed

hot press glue

nailed

electronic glue

screwed or bolted

shop

field

batts

metal foil

reflective paint

spray or brush coat

insulation backing

asphalt membrane

metal foil

Insulation

Vapor barrier

In the Wingfoot home there is a single stressed skin plywood roof
panel which also serves as a ceiling for the entire center section of
8' X 26'. ( The house also has two small bedroom extensions. ) This
ceiling is made of %" plywood, sealed, primed, and finish coated in
the plant and glued and stapled 40 to the underside of 1" X 6" and
2" X 6" ceiling-roof joists. The interior of the panels contains a
vapor barrier and 2" blanket insulation, and the exterior surface is
%" plywood.

In the Southern California Homes house, ceiling-roof panels as
large as 8' X 18' are made up of 3" plastic-impregnated paper core
to which are bonded sheets of 0.020" aluminum. Exterior joints be
tween contiguous panels are closed by an aluminum cover strip
which fits over standing edges on the panels.

A summary of the details of the various types of stressed skin ceiling
panels used by 21 prefabricators at the time of the survey is given
in Chart K.

40 Some authorities have questioned whether stapling gives enough glue-line
pressure for a good bond in stressed skin construction.

5. Solid Panels

Almost every solid ceiling panel being produced at the time of the
survey was a concrete slab used in a concrete structure, though there
were designs under development making use of Cemesto-like sand
wiches and similar materials. The concrete panels were relatively
massive and generally required specialized equipment to lift them
into place and a very strong wall structure to support them, so that
their application was necessarily somewhat limited.

In this category, however, should probably be included precast
spanning units of reinforced concrete,41 even though they ordinarily
are made with a hollow core. Such units come ordinarily about
If— 2' in width, with length and thickness varying according to the
requirements of span and loading. In place, the units are grouted
together and present a flat concrete surface, the underside of which
often serves as a ceiling. The use of prestressed concrete systems
in housing applications was also under study, but none were used for
specific prefabrication applications at the time of our survey.

All of the seven companies producing solid panel ceilings designed
them to be used as ceiling-roof or ceiling-second floor combinations.
The joints usually consisted of a lapping or keying system with a
mastic filler, although Vacuum Concrete utilized a vacuum form to
cast the joints at the site and give them high early strength. This
company was also engaged in studying the use of prestressed concrete
slabs, so designed that tensioning of the reinforcing rods in the slab
after setting would throw the concrete into continuous compression
and thereby stiffen the slab and increase its strength, perhaps as much
as three times, without increasing the amount of material used.

6. Poured at Site

The rigid structures of the LeTourneau house, the Ibec house, and
the Neff Airform house were the only ones which could be placed
in this category at the time of the survey. However, work was being
done at that time on cast-at-site ribbed concrete ceiling slabs, and on a
number of other systems which cannot properly be called prefabri
cation. Of these, among the most interesting were those which used
the floor slab as a form for the ceiling slab. In one of these, developed

41 As made by The Flexicore Co., Inc., Illinois- Wisconsin Concrete Pipe Co.,
Cities Fuel and Supply, and others.

252

simultaneously by Philip N. Youtz and Tom Slick, the reinforcement
pattern of the ceiling-roof slab was designed to permit the pouring of
this slab around properly located metal columns which contained
jacking devices. After the slab had set, the ceiling-roof could be
jacked to the proper height and fixed in position, leaving nothing
but curtain walls to be supplied to close in the house.

It was in these fields of concrete design, handling, and erection,
rather than in true prefabrication in concrete, that some of the most
interesting research and development work was being done at the
time of the survey.

F. Roofs

1. General

In most cases the roof is one of the most difficult of the house
components to prefabricate although, as a relatively simple and un
broken large area, it should lend itself to mass-production methods.
The main difficulty, of course, is the application of final roof surfac
ing, which in most cases must be done in the field to ensure weather-
tightness. Joints on the roof surface are difficult to make secure, and,
in fact, of all the elements of the house subjected to weathering, the
roofing is likely to be the least durable. The development of a good
inexpensive roofing scheme is still one of the great needs in the field
of building construction.

With regard to flat versus pitched roofs, many of the more experi
enced pref abricators agreed that it was cheaper ( estimates were given
varying from $150 to $500) to produce a flat- than a pitched-roof
structure, partly because of savings in surface area and partly because
of the elimination of gable-end walls. Yet many of these prefabri-
cators felt that it became impossibly expensive to assemble a satis
factory protective surface on a flat roof, and that pitched-roof con
struction retained several other advantages in the prefabrication of
houses, particularly if it could provide bulk storage area and space
for future expansion. In the future, the improvement of metal roof
ing skins may alter this balance. General Homes, for instance, had
designed a roof of factory-bonded aluminum skins with special joints
to be made in the field, which eliminated further field roofing work.

253

(The company was required to change to a pitched roof by FHA,
nevertheless, at a cost estimated by its chief designer at $500 per
house. )

A design idea which had been studied by several of the prefabri-
cators (Production Line, Fuller, Southern California) was that of
sealing the entire roof panel, butting panels together at the site, per
mitting some water to run down between them, and carrying it off
below in troughs on top of the structural members or batten strips
which bridge the joint. It was thought easier to maintain the in
tegrity of such joints and troughs and simpler to make the joints in
the field than in the more usual designs, and the mastics or caulking
used under the panels would be shielded from the worst effects of the
weather. On the other hand, many new and difficult problems of
freezing and clogging had to be faced before the idea could be widely
used.

2. Frame Assembly

This type of roof construction was almost invariably found in con
nection with pitched roofs where extensive use was to be made of the
attic space beneath, and where the desired continuous surface of roof
ing applied over framing members was not so easily panelized in the
factory as were other components. Furthermore, shortages of the
larger pieces of dimension lumber and planning and construction ad
vantages had led to an increasing interest in truss construction sys
tems, with some prefabricators seeking additional benefits through
the use of timber connectors. These specialized members tended to
become the subject of quantity production themselves, and accord
ingly interest was turned away from the development of roof panels
as such.

The possible advantages of the truss over the normal ceiling joist
and rafter construction, in addition to permitting factory fabrication
instead of difficult site work, lie in the use of shorter lengths of lumber,
in the use of smaller sections of lumber if the truss spacing is the same
as that of rafters or wider spacing if the sections are the same, in the
resulting clear span from wall to wall without interior bearing parti
tions, and in the design freedom resulting from this clear span.
Most truss systems break up the attic space, and so they tend to be
employed in connection with designs which make no use of this space.

254

Production advantages of the truss lie in the fact that it can be jig-
assembled by unskilled labor, that it undergoes little change of shape
after fabrication, that it is not particularly bulky to ship or handle,
and that it serves automatically to line up exterior walls and level
ceiling surface.42

In this connection it is interesting to note that several steel fabri
cating concerns and at least one aluminum fabricating concern were
offering lines of roof trusses to the general housebuilding market.

Few of the roofs in this category are insulated to any great extent
so that, in order to prevent condensation from occurring in the attic
space and to avoid overheating that space in hot weather, it is custo
mary to ventilate the roof, either through louvered openings in the
gable-end walls or through eave vents, or both, and to put a vapor
barrier or insulation layer, or both, next to the ceiling. The roof
structure thus becomes essentially an umbrella over the rest of the
house. This is particularly true in the case of some of the metal roof
systems, in which a much more serious condensation problem must be
met by employing large quantities of moving air, and to which, be
cause of their rib-like framework and thin protective skins, the um
brella analogy is much more clearly applicable.

In the Lustron house, for example, steel "hat sections" were welded
together to make up trusses which, spaced 4' o.c., supported vitreous
enameled steel pans shaped to give the general appearance of tiles
and interlocked in much the same way at the joints. Beneath this
umbrella the air was quite free to move about, while a thick blanket
of insulation protected the heating plenum chamber beneath.

The Steelcraft Manufacturing Company offered a house in which
the steel angle trusses were spaced 8' o.c., supporting a latticework
of steel angle rafters 2' o.c. and hat-section steel purlins 10" o.c. Over
this framing were laid aluminum sheets, to the underside of which
was cemented 43 Ib. felt. The edges of these sheets turned up to
form vertical standing joints. Louvered openings at gable ends were
used for ventilation.

These metal roofs were the exception, of course, rather than the
general rule. A summary of the details used by the 52 companies
using frame assembly construction in their roofs is given in Chart L.

42 For a discussion of the use of roof trusses, see HHFA Technical Bulletin,
no. 8 (January 1949), pp. 61-6.

255

Chart L

Frame Assembly Roof
52 Companies

Frame members

Structural cladding

Roofing surface

wood

steel

aluminum

1" X 6"

2"X4"

16" o.c.

24" o.c.

nailed

screwed or bolted

welded

field

shop

part of ceiling construction? no

yes

top chord of truss? yes

wood

plywood

nailed

screwed or bolted

field

shop

asphalt shingles

optional

wood shingles

field

shop

256

27 Southern California Homes house

showing

living room interior

showing garden wall and
storage unit

22 Reliance ho

23 Section of A1ROH house
being unloaded from trailer

24 Pierce Foundation— Cemesto House

25 Production Line Structures-prototype house under construction

26 Wingfoot house, showing bedroom sections extended

27 Acorn house

folded unit in place,
showing supporting
beams in place

unfolding floor
and end wall

unfolding side walls
unfolding roof
completed house

28 The Fuller house under construction
floor structure laid

mast erected and roof structure assembled
roof raised and walls suspended
ventilator hoisted on to finished structure

29 Kaiser Community Homes house

30 Green's "solar house'

3. Frame Panels

The same general considerations apply to a frame panel system as
do to a frame assembly system, and again it will be noted that the
large majority of prefabricators in this classification used wood and
made it up in a relatively conventional manner, applying the roofing
as a continuous surface at the site. In fact, since trusses were usually
fabricated as units and rarely combined in panels, frame panel roof
construction was on the whole even more conventional than frame
assembly.

There were a few prefabricators, however, who sent out panels
which were entire roofs, or large sections of roofs. These were,
naturally enough, in the group which used mass-production methods
in erection as well as in fabrication. One such was the Byrne Organi
zation, which at Harundale made up in the project shop an entire
pitched-roof frame of steel trusses to which wood sheathing and shing
ling were nailed and gable-end walls welded, also in the shop. The
complete roof was then transported to the site, placed atop the house
with a special lifting machine, and welded to the steel wall members.

The roof of the Hamill and Jones house was an unusual example
of frame panel construction in wood. In the first place, this was one
of the few hip-roof schemes. Also, the roof was made up of 4' panels
extending from eave to ridge or hip, with 2" X 4" rafters 24" o.c.
except at panel edges, where a V X 4" was used; V braces supported
these rafters at center span. The most interesting feature was the
fact that shingling was shop applied in such a way as to be woven
together over the butt panel joints at the site. This was made pos
sible by a master jig on which the whole roof assembly was put
together in the shop, and then separated into panels for transporta
tion to the site.

A summary of the construction details of the frame panel roofs
used by 54 prefabricators is contained in Chart M.

4. Stressed Skin Panels

This type of construction was seldom used for a roof unless the
underside of the panel was also to be used as a ceiling, for otherwise
the extra strength and finish were not warranted. Thus it was prin
cipally a flat-roof construction. Since it is also the lightest sort of
roof construction in most cases, special attention often had to be

257

Chart M

Frame Panel Roof
54 Companies

Frame members

Panel size

Joint

Structural cladding

wood

steel

2"X6"

2" X4"

2"X3"

2" X 8"

16" o.c.

24" o.c.

nailed

glued and nailed

screwed or bolted

welded

shop

field

part of ceiling construction? no

yes

top chord of truss ? no

yes

eave to ridge X 4'

eave to ridge X 8'

width of roof X 4'

butt

m and f

interlock

plywood

wood

nailed

glued and nailed

shop

field

258

asphalt shingles

wood shingles

optional

painted metal

built-up roof

field

shop

Roofing surface

paid to negative wind-loadings which would tend to lift the panels
free of the house.

The construction of several of these panels was described under
stressed skin panel ceilings, p. 251. Further details regarding some of
these may be given here. For the Southern California Homes house,
the addition of an asphalt-base paint over the zinc chromate primer
on the aluminum surface was considered advisable, and at the joints
there was a metal cap over the standing panel edges and a cover
strip on the underside of the panel joint to carry any seeping water
down to gutters and downspouts. The Wingfoot roof panel was
finished on the exterior with roll roofing mopped onto the complete
roof panel surface in the factory.

In the proposed Fuller house, it was planned to rivet sheets of
aluminum to W-shaped metal ribs which arched in umbrella fashion
from a central supporting mast. Moisture was relatively free to
enter the grooves of the W-shaped ribs, and in these grooves it was
to be carried down to a gutter located inside the bottom edge of the
roof. There has not been a great deal of experience with such inside
drains to test their effectiveness in the face of freezing and clogging
conditions, since for the most part they have been used in houses
designed for warm and relatively dry climates.

Most of the companies used stressed skin panels in what might be
called conventional manner, as will be seen from Chart N, which sum
marizes the construction details of the 14 companies constructing their
roofs in this way.

5. Solid Panels

In the case of each of the six companies producing them, the
solid panel roof systems were so designed that the underside could
also serve as the ceiling of the space below. The weather surface

259

Chart N

Stressed Skin Panel Roof
14 Companies

Frame members
or structural core

Joint

Structural cladding

wood

plastic-impregnated paper

steel

aluminum

V X 6"

r x4*

t*X4*

16* o.c.

48" o.c.

12" o.c.

glued and nailed

glued

screwed or bolted

shop

field

part of ceiling construction? yes

top chord of truss? no

yes

butt

special

m and f

interlock

lap

plywood

aluminum

steel

glued and nailed

hot press glue

shop

field

260

Roofing surface

built-up roll

optional

painted metal

unpainted metal

field

shop

batts

metal foil

felt

Insulation

normally was the conventional built-up roofing of felt, asphalt, and
mineral granules. A great deal of research was being devoted to the
use of homogeneous materials such as improved concretes, wood waste
composites, and rigid insulating compositions as self-supporting roof
materials which might supply finish as well as surface and which
offer at the same time the possibility of mass production of large,
simple elements. This work had not led to commercial developments
at the time of the survey.

6. Poured at Site

The few systems which fall under this heading have been de
scribed already under the same heading in the section on ceilings,
since in every case poured monolithic ceilings also served as roofs.
As a weather surface, the LeTourneau house had a coat of water
proofing applied to the standing shell. In the Neff Airform house, a
layer of waterproofing compound was sprayed on the surface just be
neath the final coat of gunite.

7. Gable-End Walls

In the many cases in which the attic was not designed for living
space, the gable-end walls were single-surface, uninsulated sections.
In all but a very few such cases, therefore, they were separate panels,
rather than a vertical continuation of the end wall construction, which
was likely to be quite different. It was also considered preferable to
make a separate gable-end wall panel so that the lower wall panels

261

could be more thoroughly standardized. Of the companies making
such panels 32 had wood louvers installed in them for ventilation,
while another 15 used metal louvers.

IV. Miscellaneous Design Features

A. Plumbing

Many prefabricators took one step beyond most of the conventional
builders in working out a standardized plumbing layout as an integral
part of the house design. Seventy-eight companies were known to
provide a standard and specific plumbing plan, detailing piping and
connections, in their blueprints. In theory, prefabricators are in a
good position to use the most advanced and economical methods
commensurate with good engineering practice; actually, preference,
prejudice, codes, the plumbing industry itself, and the unions have
made it difficult to use such rational designs, or even to use standard
ized designs of relatively conventional nature over very wide market
ing areas.

This problem was recognized by all concerned, and efforts were
being made to solve it. Particular mention should be made here of
tests made by the National Bureau of Standards, with the cooperation
of the Housing and Home Finance Agency, of private groups, and of
the plumbers themselves, to determine by scientific methods the an
swers to many long-standing plumbing controversies, and in the end
to stimulate code simplifications which will have the support of all
groups.43

As for the degree to which prefabricators are able to make use of
rational plumbing designs, it can be pointed out that 49 companies
had arranged to place their kitchen and bathroom fixtures back to
back for simplified connections from a single "plumbing wall." The
majority, however, seemed to feel that many a subcontractor refuses
to pass along the savings made possible in this way, and since to some

43 The Uniform Plumbing Code for Housing," HHFA Technical Paper, no. 6
( February 1948 ) . Preliminary edition out of print, revised edition in preparation.

262

degree the requirement that these fixtures be back to back is a limi
tation of freedom in planning of kitchens and bathrooms, it could be
disregarded.

Where the prefabricator actually put together part of his plumb
ing in the factory, however, he did not disregard the possibility of
such savings in pipe and fittings. Of the 27 who assembled their
own piping panels in the plant at the time of our survey, nearly all
used back-to-back layouts. Six other companies, which themselves
made no attempt to assemble the piping but only sent it along with
the package already precut and threaded for local assembly, also
sought back-to-back economies.

Many of the prefabricators were convinced that copper-tubing
supply lines and soldered connections were more economical than
the conventional iron, even though materials cost a bit more, be
cause the labor costs were less. Some would even have liked to offer
welded steel plumbing assemblies featuring pipe bent on tube turn
ers, but feared conflict with existing code and inspection provisions.
An interesting plumbing development was the plan of Southern Cali
fornia Homes to prefabricate its system completely in three sections,
a copper-tubing supply system, an underground waste system, and a
vent system.

The prefabrication of plumbing is one rational step which most
prefabricators were anxious to take, for plumbing is a good example
of an expensive item which can be completely standardized and
which permits mass-production and mass-procurement economies at
the same time as simpler handling and lower costs of erection. A
typical prefabricator made the point clear with a question: "A com
plete set of plumbing materials and fixtures probably costs $100, but
the installed cost of a bathroom is at least $500 or $600. Why?"

From the prefabrication of plumbing, the next step is the prefabri
cation or at least preassembly of the fixtures, too, and six companies
attached fixtures to the piping in the plant. There is nothing new
about this, or indeed about the fabrication of a specially designed
combination of fixtures in the form of a single unit. Buckminster
Fuller's bathroom design (see Figure 31) was tried out and seriously
considered for mass production a dozen years ago, and at least 10 of
the prefabricators who were visited had worked up designs for unit
bathrooms or bathroom-kitchens.

A design for a bathroom unit which was in production at the time
of the survey was that of Standard Fabrication, Inc., which had a
guaranteed market contract for 25,000 units. The unit combined

263

bathtub, adjustable shower, toilet, lavatory, storage cabinet, and
medicine cabinet with an integral low partition— all of stamped steel
finished with porcelain enamel. The final dimensions were 7%' long,

Figure 31. Patent Drawing for the Integrated Fuller Bathroom

4' high, and 2%' wide, making it possible to carry the unit through
any door and install it in any room. The plumbing was built in,
requiring only four site connections, with supply and waste lines and
vent stacks.

264

B. Mechanical Cores

The discussion of plumbing leads naturally to the mechanical or
utility core. The desire to combine all the plumbing, heating, and
mechanical elements in the house in a single centralized and mass-
produced core has long intrigued designers. Furthermore, it should
be possible so to design such a core that it can be used in a variety of
house plans, thereby combining in a single unit suited to production in
large volume many of the troublesome and costly ^elements common to
all houses within a certain range of sizes and types. Since several
prefabricators estimated that the plumbing, heating, and electrical
subcontracts alone commonly run as high as 25% of the total cost of
the house, and in frequent cases even higher, this concept was widely
studied.

A mechanical core so conceived might logically become a starting
point for the rational architectural planning of the whole house, but
few of the prefabricators felt they could afford to wait until the ideal
core was in production, or to make major alterations in their designs
to fit the cores already on the market. Still, such cores as there were
offered the advantage of very great compactness, and they were in
the main well suited to incorporation into prefabricated houses. Sev
eral of the prefabricators tried them out. Lustron used a similar
approach in the design and production of its own plumbing wall
and bathroom fixtures, and at least five others were working on their
own mechanical cores, including Fuller. The cores available on the
market, however, were not produced by prefabricators.

The best known of these was Borg- Warner Corporation's Ingersoll
Utility Unit (see Figure 32). This was a standard unit consisting of
a mechanical core plus kitchen and bathroom equipment, and con
taining the major installations, equipment, and controls for heat,
electricity, water, and gas. In the mechanical core was a forced
warm-air furnace with blower, air filter, and thermostatic controls;
an automatic water heater, either electric or gas; a prefabricated
sewer stack and vents; a prefabricated copper water-piping assembly
and gas lines; a chimney flue base with drafts or dampers, depending
on fuel; complete wiring and multibreaker for all components, plus
thinwall conduit; and a cold-air return system— all mounted within a
welded steel channel frame 30" wide, 94" long, and 77" high, with
attached wood stripping to which to fasten finishing materials. The
kitchen equipment included a 7 cu. ft. refrigerator, a single-bowl
porcelain enamel sink with supply and waste connections to the core,

265

a four-burner range, and various cabinets and lights. The bathroom
equipment included tub, lavatory, and water closet, together with
connections, standard accessories, and medicine cabinet.

Figure 32. Exploded Drawing of Ingersoll Utility Unit

The builder or contractor installing the unit had to supply plenum
chamber, ducts, registers, and chimney piping for the furnace; lead-in
lines for sewer, water, gas, oil, and electricity to the core; soffits
to the ceiling above the core; access and clean-out panels; and plaster
or wallboard sheathing and all finishing materials.

266

A "deluxe" model added laundry equipment, a few refinements, and
some rather larger items than supplied in the standard model. Borg-
Warner manufactured many of the elements in addition to assembling
the core and planned eventually to manufacture all the units con
tained in the core. At the time of the survey, the unit was generally
somewhat more expensive than comparable equipment assembled
locally, but it was sometimes able to compete because of convenience
of installation or because essential elements might not always be
locally available. A few prefabricators were using this core, but
most were proceeding with local assembly, with an eye on the rela
tive cost figures (thus making it difficult for Borg- Warner to realize
the economies possible through mass production 44 ) .

A different approach was used by Timber Structures, Inc., the pro
ducer of Mobilcore. This was actually a factory-built kitchen, bath
room, and dinette or utility room, complete and ready for attachment
to a new or old house, rather than just a core and equipment package
for insertion between kitchen and bathroom. In one model, this pre
fabricated section was 24' long, 8' wide, and 9' high, with walls, floors,
and ceiling of conventional wood frame design and floors factory
finished with linoleum. The usual fixtures, attachments, and storage
elements came in the standard model, with heater, hot-water heater,
laundry tray, storage cabinet, and exterior door in the optional utility
room, and stove, refrigerator, and mechanical washer also optional
features. When hooked up with utilities, this unit was ready to use.

An interesting sidelight of this development was the design for a
rather unusual house worked out by the company as the result of
measures taken to protect the unit during installation. In this design
the basic mechanical unit served as the load-bearing structure on
which were placed cantilevered trusses to support not only the roof
but also a set of non-load-bearing exterior curtain walls. This sup
ports the theory, often put forward, that ultimately the rational pre
fabricated house will be an outgrowth of the mass-produced mechani
cal core, rather than the reverse.

Several companies evinced interest in prefabricating complete
kitchen units. The Puraire kitchen, for example, was designed for use
in small homes and in converting large houses to housekeeping apart
ments. It came complete with sink, stove, refrigerator, and storage
facilities, and it has been used as a convenient manner of supplying
the full kitchen equipment in at least one development of prefabri-

44 Borg-Warner suspended production of the Ingersoll Utility Unit on June 30,
1949. In the words of Progressive Architecture, XXX (June 1949), 1, it "failed
to meet the present economy demand for minimum units."

267

cated houses— that for married veterans at Massachusetts Institute of
Technology. The use of or manufacture of such partial units, and of
bathroom units, has long been considered by the prefabricators but
almost none were actually used by them at the time of the survey,
for reasons of cost, design complication, and marketing difficulty.
Presumably the other problems would be more willingly faced were
the costs to come down.

C. Heating

Heating is an important element of any scheme for housing in the
greater part of the United States, and the prefabricators were well
aware of this fact. They followed with interest the recent develop
ments of the heating industry, and the new ideas on which extensive
research was being done. In order to stay within the price range
which seemed to them best suited to the concept of prefabrication,
most of them hesitated to try anything requiring large initial capital
investment. At the same time, in order to stay well within what they
considered the range of general public acceptance, most of them
also hesitated to try anything radically new.

Of the companies visited, 70 had decided upon a specific and
standard heating layout for their houses, and, of these, 50 commonly
supplied the heating unit, while nine more offered the unit at the
purchaser's option. At least 50 of the companies regularly supplied
prefabricated stacks or flues of metal or asbestos cement, and three
more supplied them at the option of the purchaser. Only a few of
the prefabricators attempted to take advantage of their quantity pro
duction to procure a heating unit specially tailored in size and char
acter to their house; the Byrne Organization designed a boiler for its
radiant-heated floor slab, and Lustron an overhead heater and plenum
chamber for its radiant heating ceiling panels.

It will be noted that, with a very few exceptions, the prefabricators
have avoided the use of hot-water or steam piping and radiator-type
heating systems, and of any system requiring coal as a fuel. This is
clearly the result of a balance of convenience and cheapness of in
stallation with convenience of operation. Certainly the most popular
heating systems from the point of view of the prefabricators were
the various warm-air systems. In areas where natural gas was avail
able at low cost, gas-fired warm-air systems were the rule.

268

1. Gravity Warm-Air Heating

Gravity warm-air heating systems were specified by 14 companies,
and, of these, 13 companies regularly furnished the furnace as part
of their package. Most popular seemed to be the floor furnace,
which takes up a minimum amount of usable living space and, with
proper design, gives reasonably satisfactory performance for a small
house. It requires no ductwork to distribute the warm air and is for
that reason probably the least expensive heater in terms of initial
cost.

In some cases, two warm-air heaters were installed, one as a stand-by
to go into action only in conditions of extreme cold. In at least three
cases, a vertical stack was added over the heater to speed up air
movement in the manner of a chimney and to get more even heat
distribution. In the Kaiser Community Homes house, such a stack
had louvered openings near the ceiling, which, combined with the
cool-air intake near the floor, provided a better circulation of air
than is usually possible in basementless gravity units.

2. Forced Warm-Air Heating

Forced warm-air heating was specified by 22 companies, although
in several cases only as an optional feature. Only a few found it
advisable to furnish such units for use with conventional duct sys
tems. At least 13 companies supplied prefabricated ducts, plenums,
and even risers with their packages, and, when required to do so for
houses with basements, they provided the necessary risers, ducts, and
grilles in the walls. In basementless houses, some of the companies
supplied prefabricated plenum chambers of metal or asbestos cement,
usually concealed above a dropped ceiling in the central hall, to
distribute the heat from the warm-air discharge to various sections
of the house.

Interesting technical developments coming on the market at the
time of the survey but not yet used by the prefabricators included
very compact furnaces designed to make use of high-powered com
bustion principles worked out during the war for aircraft heating and
to distribute the heat through flexible hosing instead of ducts; these
were based on concepts of few moving parts and high operating
efficiency.

269

3. Radiant Panel Heating

Widely discussed in general, radiant panel heating systems have
been of particular interest to the prefabricates. Since such systems
require careful engineering and close adjustment to the characteristics
of the house in which they are to be installed, and since the cost of
installation in some cases shows promise of coming down to the
small-house range only through the application of industrial tech
niques and repetitive production, their use would seem to offer a spe
cial advantage to those who manufacture a limited range of models
in large-quantity production and who can, therefore, afford to spend
money for a careful integration of the house and heating design and
for the necessary tooling up. Houses which offered radiant heating
without excessive cost, whether in the floor, walls, or ceiling, and
whether by means of warm air, hot water, or electric resistance ele
ments, appeared to have gained, because of general public interest
and favor, a sales advantage over those which offered conventional
heating.

Radiant floors were the most common of the radiant systems in
stalled or specified by prefabricators at the time of our survey. At
least nine of the companies were using them or were planning to do
so, six with concrete floor slabs in basementless houses and three with
wood floor systems. Of the concrete slab installations, the majority
circulated hot water in copper tubing. The Harundale project of the
Byrne Organization had a 40-gallon, specially designed hot-water
boiler which also supplied the hot-water needs of the house, and a
pump to circulate the hot water through copper tubing which was
spaced at V intervals in most of the slab and somewhat closer in the
kitchen and bathroom. In this fairly typical installation, slab surface
temperature was designed for 85° F., and it was hoped to keep heat
loss into the ground as low as 10%. The economies of combining do
mestic hot water and heating, and of obtaining an efficient design
of floor slab for a cold climate, together with potential fuel economies,
had to be balanced against somewhat higher installation cost and
several recognized design problems. These included the problems
caused by time lag in supplying or in reducing heat because of the
large mass of the concrete slab, the need to provide against accidental
blocking or progressive obstruction of the tubing, the complications
caused by rugs and other floor coverings, the interferences resulting
from other plumbing and mechanical installations, and the measures
which must be taken to protect against freezing the water and burst
ing the tubing during unheated periods in extremely cold weather.

270

Three companies had erected experimental models of systems which
made a radiant panel of a wood floor system by using the space be
tween the floor and the ground slab of a basementless house as a
plenum chamber heated by forced warm air. In an experimental
house built by the Field Detroit Co., warm air was forced into the
plenum at the center and returned to the house through registers
located beneath the windows. In the Green's Ready-Built house,
warm air was first circulated through the house, then returned through
registers to the underfloor plenum, and eventually carried back to the
furnace. (This system was not approved by the FHA.) The Solar
Homes Co. used warm air circulated in the space between the floor
joists, which was covered and insulated at the bottom to produce
a sort of duct.

Three other companies were experimenting with forced warm air
in various hollow slab or tile systems.

Radiant walls were not in production among the prefabricators
visited, although a few were experimenting with them. This was in
part the result of the difficulty of designing a heating system for uni
form performance at different distances from the walls and for sur
faces which are complicated by numerous openings and many insula
tion problems.

Radiant ceilings were scheduled for regular production in only two
houses. One of these was the Lustron house, the heating system of
which has already been described on p. 247. The other was the Mod
ern Standardized Buildings Co. house, in which warm air was heated
as it rose through a series of finned hot-water coils and was then
conducted into a plenum at the ceiling, made up of a sheet of %"
asbestos cement which was suspended 2" below a sheet of aluminum
foil backed by 2" of blanket insulation. A %" open space between
the edge of this plenum and the wall permitted the warm air to
circulate down into the house area and then to return to the heater.

These two systems reflect the design interest in ceiling installations
which many consider the most satisfactory from the point of view
of the physiological needs of the body. Unhampered by the large
mass of the floor slab and by the problem of floor coverings, the ceil
ing panel can be highly efficient in its action and can greatly assist
the development of open and flexible planning; nevertheless, it cre
ates the new difficulties of installing a somewhat complex structure
at ceiling height and of insulating it against heat loss upward into
the attic or roof areas. Other new techniques such as the use of
electric resistance panels, either of a special rubber material such as

271

the United States Rubber Company's Uskon, or of conductive alumi
num alloy fused into tempered glass, were being carefully studied
where costs of electric power were sufficiently low (1%^ Per kilowatt
hour or less), because of the possibilities offered for extremely low
installation costs and efficient use of heat.

4. Solar Heating

Very few companies at the time of our survey were attempting any
considerable use of solar radiation to heat their houses, and those few
were interested only in the use of large so-called "solar" windows,
facing directly south and so shielded that the sun can penetrate the
window and warm the interior space in winter but cannot penetrate
in summer. This idea has captured the imagination of many archi
tects and, to some extent, the public. The technical, psychological,
and esthetic arguments and pronouncements pro and con are well
known and need not be rehearsed here.

Green's Ready-Built solar house was designed expressly to utilize
solar heat. Most of the living spaces had large windows to the south,
so that the house tended to be long and narrow, running from east to
west. The design of the windows, with fixed glass and openable,
louvered, and screened ventilation areas above and below, has been
discussed on p. 232. There was also added to the exterior of the house
a series of louvered screens projecting from the wall between window
areas, which served to protect the windows against late afternoon sun
shine without interfering with the flow of air along the southern front
of the house. This house, frankly designed for the middle-income
group, received a good deal of architectural acclaim, but it was never
put into mass production, at first because of limitations imposed by
the Veterans' Emergency Housing Program and later because of the
failure of the company.

No prefabricator was known to have made any attempt to put into
production a heating system designed to trap the solar radiation im
pinging on the house, to use some of it for immediate heating, and to
store the rest for later use when solar radiation is not available. Re
search in this field is followed with interest in many quarters, but is
not yet at a stage where its advantages or disadvantages are clearly
known.

272

D. Electrical Wiring and Fixtures

The intended position of electrical outlets was specified by 69 com
panies, and 49 prefabricated their wiring to some degree. Of these
latter, 11 simply precut it, nine factory-installed dummy lines in the
conduit to help the electricians pull through the wires in the field, and
27 actually preinstalled the wiring, usually in wall panels. Most of
these companies used BX or Romex cable rather than rigid metal
conduit, and in some communities they encountered code difficulties
on that score.

The average prefabricated wiring system had outlets in the wall
panels, with the connecting cable stapled or clipped by brackets to the
structural members or run through conduit cast in the walls. The
connecting wire usually projected through the top or bottom edge of
the panel, with the balance of the wiring placed inside during ship
ment, ready to be pulled into position in the attic or crawl space at
the site, and attached to major circuits and to service leads from
outside.

A few companies had devised systems for prefabricating the major
circuit wiring (for example, by stapling connecting wires to a board
running the length of the attic space) so that all that was necessary
at the site, after connecting house circuits, was to connect this "har
ness" to the service leads. In such a case, the meter and a master
circuit breaker or fuse box, or both, were also wired and installed on a
panel in advance.

Most prefabricators felt that the savings possible from prefabricat
ing their electric wiring, if their construction permitted it at all, were
not great enough to justify the probable difficulties with code and
labor which would result. Many companies which had no wall panels
adapted to prefabricated wiring of this sort used a channel behind
the baseboard or a molding attached to the baseboard to carry the
wiring inside it. In concrete systems there frequently were metal or
cardboard conduits cast in place. In all systems there seemed to be a
general desire to keep the outlets at or near baseboard level, and to
run cable for ceiling or wall fixtures or switches up behind batten
strips at joints, or inside door casings. For example, Southern Cali
fornia Homes used a 2" box section at the base of its panel, among
other reasons, for wiring, and ran wiring for three wall and overhead
fixtures and a few switches inside the channels which served as the
edges of the panels, the wiring being placed in these channels before
they were bonded to the skin surfaces of the panels.

273

Few companies had given special attention to the design of lighting
fixtures or to a general lighting plan, although in this field lies another
possible advantage of the prefabricator over the conventional builder.
Such companies as were concerned about fixtures had for the most
part concentrated on getting maximum procurement advantages from
mass orders without, so far as was seen, attempting to engineer and
produce a house in which lighting might reach a standard of high
performance.

E. Acoustical Treatment

A very few companies had given serious attention to the acoustical
properties of their houses, and these were the companies offering
houses of unconventional design or construction which might give
rise to particular problems. Even they relied primarily on the
furnishings for the deadening of sound. Only one house (Green's
Ready-Built) is known to have made use of acoustical tile for its
standard ceiling surface.

No attempt was made in this survey to make scientific acoustic
measurements or tests, but there were many houses in which reverbe
ration was noted, and several more in which the partitions and doors
could not have been much above the level of acoustical transparency.
The problem would have been much more prominent had the com
panies visited not limited their interests primarily to the prefabrica-
tion of small detached houses. No doubt, more attention will be
devoted in the future than at present to the noise problem, even
in such detached houses, for in lightweight construction noise may
cause a great deal of unpleasantness and yet its solution is known to be
neither complex nor unduly expensive.

F. Built-in Furniture

Many of the prefabricators were aware of the advantages they could
offer over conventional builders by building in elements of furniture
and storage at the plant, particularly when it was possible in this way
to make use of scraps of material which would otherwise be wasted
and, by the use of their regular equipment, to produce articles which
could hardly be duplicated at the site for comparable costs. Thus

274

one company (Red-E-Bilt Homes) offered a series of inexpensive
built-in features (a mahogany- veneered ceiling in the entry, a kitchen
ceiling vent, ironing board, and spice cabinet, and, in the garage, a
workbench) which, in its opinion, added considerably to the sales
appeal of the house.

At least 73 companies regularly supplied kitchen sink cabinets as a
part of the package, and 66 supplied other kitchen cabinets, but six
companies supplied kitchen cabinets only as an optional feature. All
their cabinets were manufactured by 31 companies, in each case out
of wood. Several said they did so only because of cabinet shortages,
but many others found that they could in this way make use of facili
ties which were well suited to this kind of work.

Other built-in features included dining tables (seven companies),
bureaus, or dressing tables and drawer space (eight companies), and
storage or closet walls (17 companies). All these features appeared
to be growing in popularity, particularly the use of storage or closet
walls in place of partitions, with the storage and closet areas prop
erly subdivided and drawers provided where necessary. Such storage-
walls were supplied primarily in construction systems permitting the
use of non-load-bearing partitions.

Building in furniture to a greater degree than this was limited
to a few companies offering minimum plans in which high efficiency
was possible only through special design of the furnishings as well
as of the structure itself, and to a few other companies offering houses
for special purposes; for example, in the construction areas of the
Tennessee Valley Authority, where it was found practical and eco
nomical to build completely finished houses, to transport them to the
site in sections, and to move them away to a new site when the occa
sion demanded.

Though none of the prefabricators had actually made designs, a
few were considering plans in which certain standard elements such
as beds might be built in, which would afford the extra saving of not
having to finish the floor under the bed as well as the possibility of
using the space under the bed for storage drawers.

G. Space Arrangement

Without a detailed consideration of floor plans and other architec
tural elements, the quality of architectural space planning cannot be
properly discussed. On the other hand, space planning in small

275

Figure 33. Plans of Selected Prefabricated Houses. These plans are presented on
a uniform 4' grid so that the allocation of space and the overall areas may readily
be compared. Kitchen, bath, and utility area have been shaded. A number of
unconventional houses have been included for contrast, as have the plans of

three-bedroom models.

276

Gunnison Homes • Basic

Shelter Industries

Ibec Corp.

Lustron Corp. - Standard

Southern Calif. Homes

Crawford Corp. • Basic

Lustron Corp. - Economy

American Houses • Basic

Airform House • Typical

Fuller House

BR.

LIVING

LeTourneau • Typical

Green's Ready-Built Homes
Veterans

Lustron Corp.

Gunnison Homes

BR.

BR. BR.

BR. HE 3 DINING

LIVING

Crawford Corp.

PORCH

BR.

LIVING

houses is largely dictated by the needs of family living, and while
there is no single rational approach to the small house plan, the com
bined pressure of needs and costs greatly limits the choices open to
the designer. Most of the plans were in many respects similar, as a
result (see Figure 33). This aspect of the subject has been well
treated in architectural sources and needs no further development
here.

It is important, however, to give some idea of the standards of space
provided by the pref abricators at the time of the survey. The follow
ing summary averages the room sizes of the most popular plans and
models of typical companies. The figures give the number of com
panies from which the average was compiled as well as the average
itself:

Number

Average

of Com

Size

panies

Room

Of.//.)

Living

199

Living-dining

220

Kitchen

Kitchen-dining

116

Dining area l

Master bedroom

136

2nd bedroom

108

3rd bedroom 2

113

Closet area

Other storage area

Utility

Covered porch

1 There were very few companies with separate dining rooms. These were usually ells.

2 More two-bedroom than three-bedroom houses were minimum in space planning.

Combining the figures for average room sizes, we find that the three
different combinations of living-dining-kitchen areas result in aver
age houses (allowing 10% wastage in floor planning) of the following
floor areas: two-bedroom house, 770, 773, or 830 sq. ft.; and three-
bedroom house, 894, 898, or 955 sq. ft.

Trends in the matter of space planning included an increasing use
of truss roofs, with the resulting freedom of interior arrangement, and
an increasing interest in the three-bedroom house. The Kaiser Com
munity Homes project in Los Angeles built only three-bedroom houses,
and many of the more conventional wood prefabricators were turn
ing to story-and-a-half houses in order to get extra bedroom space,
although at the time of the survey many of them were leaving this
second-story space unfinished for later finishing by the purchaser.

280

The prefabricated house, on the average, was not generally dis
tinguishable from the conventional product in respect to the space
allocated to closets, utility room, bathroom, or kitchen, although there
were instances in which prefabricated houses offered increased effi
ciency in respect to all of these. Of the two, there was more likeli
hood that one would find space wasted or inefficiently used in the
conventional house. In a carefully prepared ratio of usable or effi
cient space to price, the average prefabricator would probably do
better than the average conventional builder.

The previous summary indicates that at least 13 companies offered
covered porches. Despite many architectural and marketing advan
tages favoring such a combined entranceway, play, recreation, and
storage space, it was not generally offered as a standard part of the
prefabrication package.

There had been little experience with the practical effectiveness of
the very different space relationships offered in houses of specialized
construction and form, such as the Fuller or Neff house, since almost
no houses of this type had reached the common market in this coun
try. Published descriptions of these houses met with a strong interest
and, in certain segments of the public, with a decided approval.
Further than this, the many attractive adaptations of the "Quonset"
hut structures as houses in the last few years have added to the evi
dence that space planning and relationships quite different from those
believed to be required by the general public may be found to be
entirely acceptable.

H. Product Variety

1. Quality Standards

More than one quality standard in their houses were offered by 15
companies, with the difference lying primarily in the materials used,
in the size and spacing of members, or in the degree of finish and
number of fixtures added. A common practice was for a company to
offer one house for sale on the general market to individual homeown
ers, and substantially the same house, but of different quality stand
ard, for sale to industrial concerns or developers buying groups of
houses for rental or sale, typically in new communities. Or a com-

281

pany might offer one version of a house which would be eligible for
FHA insurance or pass certain building codes, and another version
which would not.

Typifying the latter production plan was U. S. Homes, Inc., which
offered a "Suburban Home" and a "Village Home." Essentially the
same in general appearance and floor plan, the "Suburban" had a
finish floor of oak and inlaid linoleum, while the "Village" had south
ern pine tongue and groove flooring and printed felt-base linoleum.
The "Suburban" had bathtub, laundry tray, medicine cabinet, hot-
water heater, and forced air heater with ducts supplied with the
package, while the "Village" had none of these. The "Suburban" also
had linoleum covering over its sink cabinet.

2. Basic Design Standards

Several basic design standards were offered by 13 of the companies,
either to reach a wider market or to experiment with very different
designs produced at the same time and under the same conditions.
Thus, Texas Housing Co. produced a "Town and Country" house of
768 sq. ft. floor area with more or less conventional wood frame
design, and also a 16' X 16' "Homette" developed from army hut
ment designs and having a pyramidal roof of the same 2" plywood
panel construction as the walls. The former was acceptable for FHA
insurance and would pass most building codes, while the latter was
designed for emergency situations, for public or university temporary
minimum housing, or for owner-built houses for temporary use or
in minimum housing areas.

3. Architectural Styles

The houses of any one company were usually of a single architec
tural style or character, but 17 companies offered two styles, four
offered three, and five offered more than three, most of the last be
ing those companies which were prepared to make up nearly any
style desired on a job-lot basis. Often those offering more than one
style were offering a "modern" house with flat roof to see how well it
would sell in the prevailing market, and generally the companies do-

282

ing this reported that their "modern" models were not doing very well.
This may represent a market prejudice, but it was probably also a
reflection of the quality of the design which some of the prefabri-
cators termed "modern."

There has been an earlier discussion of the general matter of archi
tectural style on pp. 194-6, but it may be added that most of the prefab
ricated houses on the market at the time of the survey were supposedly
reminiscent of, and often termed, Cape Cod cottages. Since many of
the prefabricators preferred not to have to supply such details as
shutters (where these were supplied they often were long, hanging
well below the window sill to give an appearance of larger windows ) ,
colonial entrances, and the like, there was also a tendency to call the
resulting unornamented, pitched-roof house by the style name of
"American cottage." There was also a tendency towards a new
name, and partly also towards a real style, called the "ranch house,"
theoretically all on one floor, with large windows and good cross
ventilation, rambling in character, and incorporating several outdoor
living areas, although in a small prefabricated house these aims were
rarely accomplished in fact.

4. Achieving Variety

There appeared to be, at the time of the survey, a definite opinion
on the part of many prefabricators (and probably of much of the
general public as well) that some sort of variety would have to be
offered in prefabricated houses, on the theory that people insist on
individuality in their homes, particularly if large projects are to be
made up of the same units. However, the architects of some of the
companies were satisfied that variation in actual form and space ar
rangement could be very limited and still give the public what it
wants; they tended to agree that the worst form of monotony is the
monotony of slight variation. Some of the largest enterprises were
offering a single model, without even the variation of left and right
plans, presumably on the theory that mass production should start
with maximum standardization and that the resulting product should
be superior enough to that of its competitors to attract purchasers de
spite any emotional resistance they might feel at first to the idea of
standardized houses. The subtleties and merits of the arguments pro
and con need not be rehearsed here, but it does seem to be a fact

283

that the largest volumes, and in many ways the lowest costs, were
being reached by those offering the least variety and relying on color
and good site planning for the creation of distinction within a general
pattern.

Many more prefabricators, however, sought to make nominal vari
ety possible on easy terms by offering different types of entrance, of
exterior finishing material, of window trim, of shutters, porches, gar
ages, breezeways, and of colors, all amplified by right and left plans.
Of these, a substantial number found it practical in effect to limit
their prefabrication to a chassis of some sort and to finish the houses in
the field with a wide variety of "treatments." They represented a
sort of mid-position between those who believed that maximum econ
omies could be achieved only by strict standardization of the com
plete house and those who believed that the public could best be
served by a maximum standardization only of the panels and assembly
parts, leaving it to the local man to put these elements together in
any form he might wish. It has not been possible as yet to determine
with any degree of assurance what the market will show with refer
ence to these varying points of view.

5. Models

Assuming the same structural system and architectural style, many
companies offered variety through different floor plans— sometimes
different only in arrangement, but usually different also in size. At
the time of the survey, 14 companies offered two plans, seven offered
three plans, 12 offered four or five plans, 16 offered five to ten plans,
five offered 11 to 25 plans, two more than 25 plans, and six any plan
at all, depending upon the order. Further than that, at least 21 com
panies offered both right and left plans.

Several of the companies, after having offered a variety of models
and having learned that certain ones sold badly, discontinued produc
tion of the less popular models. Thus, Kaiser Community Homes
originally offered both two-bedroom and three-bedroom houses in
the Los Angeles area, but at the time of the survey only the three-
bedroom model was being produced. Harnischfeger, after having
limited the number of models, was able to achieve further economies
by producing the components of the remaining models in larger panel
sizes.

284

6. Flexibility within the House

There is a common belief that one advantage of the prefabricated
house is the ability to alter or add to it at pleasure. Actually only 22
companies emphasized the possibility of adding rooms or wings, sup
plied by them, at a later date. Few, if any, stressed the possibility
of moving partitions about to meet changing family needs or de
veloped designs pointing out various alternatives, even when truss
roof construction would permit this. Although certain designs, par
ticularly the modular panel systems with universal interlocking joints,
had a basic demountability which might lend itself to rearrangement
or addition over the years, particularly if this desire is taken into
account in planning the layout and capacities of the mechanical sys
tem, there had been little actual experience with this sort of thing
other than the experience with the demountable war housing. When
the time came to move these houses, all of which were carefully
designed to be demountable, most of them were simply sawed into sec
tions and carried off without taking the trouble to follow the dis
mantling system. Nevertheless, many of the prefabricators offered
schemes which were to some degree demountable and which offered
possibilities of obtaining elasticity of plan through moving component
elements, adding other elements, or taking elements away and replac
ing them with others. Important economies might be gained through
planning for re-use value or resale and secondhand value. None of
these possibilities had been tested in the market, however, and the
tremendous site planning problems involved had hardly been con
sidered.

285

Part A A •

Chapter C_7

PROCUREMENT

This chapter and the following one are devoted to two phases of
the actual manufacturing process: procurement and production. The
distinction between these two seems easy to make; procurement is the
purchasing of the materials and finished goods which pass through
the prefabricated plant, and production is the actual business of
performing work on material to produce a product. Yet the line is
sometimes hard to draw; for instance, are those prefabricators who do
no more than assemble a house package from fabricated components
purchased from others engaging in procurement or production? This
difficulty of distinction emphasizes the point that economies may ac
crue to the prefabricator as much because of the large scale of his
operations, particularly in procurement, as because of his skill in fac
tory production per se. It is important to understand the advantages,
real or potential, which stem from each of these two phases of the
manufacturing process.

Procurement advantages stem primarily from size. It is the fact
that the prefabricator buys in carload lots, that he must spend large
amounts of working capital for materials inventories, and that he is in
a position to carry out himself the functions of handling, sorting,
grading, and repackaging, which puts him in a position to buy direct
from the manufacturer. Such advantages are, naturally, available to
either a prefabricator or a large operative builder, and it is a fact that
at the time of the survey the largest operative builder and the
largest prefabricators were of roughly the same size, judged by their
volume (1,000-3,000 houses per year).1

In the following pages are examined the prefabricator's procure
ment policies and the extent to which he was, in fact, able to lower
his costs in three kinds of purchases: raw materials, finished items
for which he acted as a jobber, and fabricated components.

I. Raw Materials

It is significant that most companies at the time of the survey
thought that cost reductions were more likely to be realized through
a change in the conventional materials distribution system than

i Levitt, 3,000 in 1947; American Houses, 1,600 in 1947; Kaiser Community
Homes, 2,500 in 1947; National Homes, 2,500 in 1947.

289

through any other single factor in the whole realm of manufacturing.
This was particularly true, of course, of those companies which
counted on volume production rather than on unconventional con
struction or materials for their economies, and it seemed to be the
case whether or not the firms had as yet actually realized such econ
omies.

Certainly there would seem to be potentialities for cost reduction
in the distribution of materials. The following table shows one esti
mate of the cost of distributing one dollar's worth of several building
materials from the manufacturer to the site:

Cost of Distributing Building Materials through Conventional
Channels

Cost of Distribution

Combined Profits

Material

Cost of
Manu
facture

Manu
facturer

Jobber
or
Whole

Re
tailer

Cost of
Trans
porta
tion

Manu
facturer

Jobber
or
Whole

Re
tailer

Deliv
ered
Price
at Site

saler

Lumber

$1.00

0.23

0.33

0.54

0.34

0.21

0.09

2.83

Plaster, lath, and

wallboard

1.00

.28

.89

.35

.53

....

.21

3.26

Insulation

1.00

.37

.09

.27

.18

.09

2.18

Concrete and

mortar

1.00

.32

....

.18

.19

.17

.09

1.95

Source: Housing Costs; Where the Housing Dollar Goes, National Housing Bulletin 2 (Washington: National
Housing Agency, October 1944), p. 46. The data on which this breakdown is based were obtained from un
published studies of the Office of Price Administration and in general represent 1940-1941 conditions.

There are obviously inviting targets here for any sort of large house
building enterprise, and this is one point where the largest operative
builder, along with the largest prefabricator, has had a degree of
success.2 It is difficult to define the degree to which many prefabri-
cators succeeded, because at the time of the survey they were not able
to buy under what might be called "normal market conditions." It
may be said, however, that only a few companies were able to effect
really substantial economies in their materials purchases, and almost
none had effected all those which they thought they should have.
Even so, it is true that the industry's dozen largest members had, for

2 See, for instance, the talk by William Levitt to The Producers' Council, Oc
tober 1947, reported in The Architectural Forum, 87 (November 1947), 10, and
in Journal of the American Institute of Architects, IX (June 1948), 253-6; and
the rebuttal letter by H. M. Long in The Architectural Forum, 88 (January
1948), 18. See also the builders' and manufacturers' points of view as pre
sented in testimony before the Joint Committee on Housing of the 80th Congress,
High Cost of Housing (Washington, 1948), pp. 127-30.

290

the most part, been able to buy directly from the mills, or in some
cases where they had been frustrated in their attempts to do so, they
had begun to manufacture materials themselves, acquiring lumber
mills and even engaging in logging operations. Most of the small
manufacturers, on the other hand, were not established in volume or
reputation and were finding it very difficult to secure accounts with
direct sources. During the Veterans' Emergency Housing Program
it was expected that the prefabricators would be able to buy directly
from the mills, but after the program ended a number of mills refused
to sell directly any longer, with unfortunate results for some of the
smaller companies. This refusal is not difficult to understand. Under
the VEHP allocation program of 1946-1947, 300,000,000 sq. ft. of ply
wood was shipped to prefabricators during a one-year period at the
direction of the government.8 The mills did not believe that the
prefabricators were able to use more than % of that amount,4 and
some sources 5 put the figure as low as %. By late 1947 and during
the first half of 1948, the situation had changed substantially, and
many prefabricators reported difficulty in securing sufficient plywood
at "reasonable" prices.

It is estimated that half of the 37,400 prefabricated houses pro
duced in 1947 were constructed chiefly of plywood. The industry's
consumption of this material in recent years is shown in the following
table:

Estimated Consumption of Softwood Plywood, 1946-1948

(millions of square feet, %" equivalent)

1946

1947

1948

Esti

Per

Esti

Per

Antici

Per

mated

Cent

mated

Cent

pated

Cent

Con

sumption

Total

sumption

Total

sumption

Total

Prefabricated houses

5.4

100

6.3

150

8.3

Other residential construction

459

35.8

643

40.6

731

40.6

Source: U. S. Department of Commerce, Construction Division, Construction and Con
struction Materials; Industry Report (April 1948), p. 7.

3 Source: Douglas Fir Plywood Association, in a letter to the Bemis Foundation,
August 21, 1948.

4 Loc. cit.

5 Construction and Construction Materials; Industry Report, U. S. Department
of Commerce, Construction Division (April 1948), p. 6.

291

Corporate integration reaching back to the raw materials stage was
observed in quite a few instances, particularly among those firms
working with wood. At least 14 prefabricators were cutting their
own timber or were closely affiliated with logging operations, while
38 companies were known to have equipment for various aspects of
lumber manufacture. Two of the companies producing houses or
house components (General Plywood Corporation and Buffelen
Lumber and Manufacturing Co.) were also producing plywood, and
others were trading peeler logs cut on their timber stands for the
final plywood product. Other examples of this type of integration
include the purchase of a sheet-steel mill by Borg- Warner for the pro
duction of its Ingersoll Utility Unit, and such tie-ups as that between
Precision-Built and Homasote or between Lustron and Chicago Vitre
ous Enamel.6

Another sort of integration which at least five prefabricators utilized
in the procurement of materials was the establishment of subsidiary
wholesale lumber and supply companies, which made it possible for
them to buy material and equipments at the lowest possible prices.7
Many of the companies, as has been pointed out in the chapter on
management, started as, or were backed by, lumber companies and
so were able to enjoy this advantage without the necessity of setting
up separate purchasing entities. At least 30 companies were in this
category. As might be expected, some of these subsidiary purchasing
organizations found it expedient to sell on the open market as well.
Not only did they sell materials which they had bought in excess quan
tities, but they also served as outlets for items such as doors, windows,
or cabinets which their parent companies might have produced in ex
cess in conjunction with their house packages.

This sort of subsidiary purchasing organization was established
by Ivon R. Ford, Inc., after the war for its own use, and that of
nine licensees in various parts of the country, under the direction
of Guy C. McKinney in Washington, D. C. In 1947 this purchas
ing subsidiary became an independent organization, and its services
were made available to the entire membership of PHMI. Purchasing
powers were pooled, and McKinney & Co. was able to arrange with
materials and equipment producers for large orders on a steady basis.

6 Precision-Built was a subsidiary of Homasote Company and made extensive
use of its materials. Similarly, the Lustron Corporation was in its early days a
subsidiary of Chicago Vitreous Enamel Product Co.

7 Texas Housing Co., California Prefab Corp., Ivon R. Ford, Inc., Brady Con
struction Co., and Claude T. Lindsay, Inc. The savings made by Levitt on Long
Island through such a device have been well publicized.

292

It was hoped that these producers might become, in effect, "perma
nent suppliers" of the service, which would receive its income from
a monthly fee paid by participating companies. Although such a
service might have been a boon, especially to smaller operators whose
activities had been severely curtailed before participation by inability
to purchase sufficient materials at reasonable prices in that confusing
period, it did not work as planned. During its existence the service
paid mill prices or slightly more for the materials and equipment it
bought but, in most cases, less than the wholesale price. Dimension
lumber, for instance, was furnished at less than wholesale quotations,
and kits of electrical fixtures were bought at savings of 35%, 40%, and
even 50% over local wholesalers' prices.

In 1946 and 1947 there was considerable criticism on the part of
the materials producers that too many prefabricators had placed orders
based on huge estimates of production and that, when these estimates
were not fulfilled, cancellations came thick and fast. Although confi
dence in the prefabricator is not the only condition that must obtain
before a materials producer will be willing and anxious to sell direct
to him, it is one of the most important ones, and the prefabricators
were anxious to achieve it.

II. Finished Material and Equipment

Apart from the structural shell fabricated by the house manufac
turer there are various components which he seldom if ever fabricates,
but rather buys in large quantities to furnish with the house package.
There are several reasons for his doing so. One is the possibility of
supplying these items to the ultimate consumer at costs lower than
those the consumer would have to pay; generally speaking, such econ
omies as were obtained in this way were very modest, although in
several instances they were considerable. Another reason is the ad
vantage gained from marketing as complete a package as possible.
Most of the advantages accrue directly to the dealer-erector, however,
and only indirectly to the prefabricator. They include, for instance,
the time saved the dealer in procuring his materials, and the elimi
nation of wastage at the site through the use of the proper amounts

293

of material. Such advantages were particularly important during
the period of the survey, which was one marked by frequent short
ages and irregularities in materials flow. The prefabricator was not
always in a position to help his dealers in this way, but when he was,
the savings obtained constituted real economies to the dealer and
probably more than offset the prefabricated storage and handling
costs. Thus costs were lowered even if purchasing was not done at
particularly advantageous prices.8

A further reason why the prefabricator may strive, by acting as a
jobber, to furnish as complete a package as possible is that by so do
ing he enables his dealer to make a larger dollar sale and thus to ob
tain a larger profit. Of course, there is also the opportunity for the
prefabricator to take a substantial middleman's profit, and many
did so.

Number of Companies Known to be Acting as Jobbers for Various
Finished Components

Regular Part

Item of Package Optional

Electrical fixtures 26 6

Flooring 58 16

Furnaces 50 9

Heating stacks 25 3

Hot-water heaters 41 8

Kitchen cabinets 73 6

Plumbing fixtures 31 12

Refrigerators 10 5

Roofing 70 4

Screens 43 12

Stoves 13 7

The table above gives the number of companies known to be acting
as jobbers for various finished components. That there were not
more companies furnishing plumbing assemblies and electrical fix
tures was largely due to the fact that prefabricators wished to make
it possible for their dealers to subcontract plumbing and electrical
work. Such equipment is customarily sold through the contractor
who installs it, and most contractors were understandably reluctant
to install fixtures included in a prefabricator's package when in so do
ing they would lose their selling profit. Since these contractors had
plenty of regular business, and since the prefabricator was often new

8 Note that savings through simplified purchasing and elimination of waste
have also been the objectives of the "industry-engineered house" program and
the proposed program of the Research Institute for Economic Housing, New York
City, Spring, 1948.

294

to them, it was frequently necessary for the prefabricator to eliminate
certain items from his package in order that his dealer-erectors might
establish successful working relationships with the local contractors.

A second factor that sometimes entered the picture was the refusal
of some local unions to handle such items as preglazed sash and pre-
hung doors. Few prefabricators, however, mentioned this as a major
difficulty. Another factor that affected the activities of some prefabri
cators as jobbers was the choice of marketing pattern. One firm
which was planning to sell through department stores said that it
would have been able to furnish refrigerators with its house packages
at 60% of the retail price, but that it could not do so for reasons con
nected with the merchandising policies of its outlet.

Considered as a whole, prefabricators were not achieving sub
stantial economies by acting as jobbers, even though they may have
bought in carload lots. Thus, much the same situation obtained in
their procurement of finished materials and equipment as in their
procurement of raw materials. Generally speaking, the "average
prefabricator" in his role as a middleman was able to offer his dealer
prices which were the same as or only slightly lower than would have
been paid to the regular distributive outlets. The mere fact that a
prefabricator might be classified by some as a manufacturer rather
than as a builder did not, it seems, entitle him to a special discount.
This was particularly true after the Veterans' Emergency Housing
Program ended when, as in the case of raw materials, some producers
of home equipment refused to continue selling to house manufacturers
at factory or even wholesale prices. One prefabricator who had been
furnishing along with his house a certain shower unit that retailed
for $65 had, during the VEHP, been able to buy the unit direct from
the manufacturer for $21, and supply it to his customers for $35, in
stalled. After the expiration of the program, the manufacturer de
clined to sell the unit in any way except through the intermediary of
wholesale and retail plumbing houses.

There were a number of cases that differed markedly from the
"average"— cases in which prefabricators had sufficient volume of suffi
cient power to obtain real savings in their purchasing. One of the
largest firms in the industry stated that its experience showed net
savings in finished materials and equipment costs of about 35% over
small conventional builders. Another company which did some
what less business, but was affiliated with a very large materials pro
ducer, reported that it was buying its jobbed materials at industrial
discounts and selling them to its dealers at a 15% mark-up. Hot-water
heaters which retailed at $150 were bought by this company at $44.

295

An eastern company stated that it had been quoted unit prices on a
set of kitchen cabinets of $48 per set. If, however, it had ordered
10,000 units, the price would then have been only $17 per set. The
same firm estimated a saving of $1,000 per house through such mass
purchasing if production volume were raised from 100 to 21,000
houses per year. A study of the proportion of the final cost of finished
materials and equipment that is represented by distribution costs
confirms the opportunities afforded for savings by direct purchasing:

Cost of Distribution of Finished Materials and Equipments through
Conventional Channels

Cost of Distribution

Combined Profits

Finished
Material
and
Equipment

Cost of
Manu
facture

Manu
facturer

Jobber
or
Whole

Re
tailer

Cost of
Trans
porta
tion

Manu
facturer

Jobber
or
Whole

Re
tailer

Deliv
ered
Price
at Site

saler

Finish hardware

$1.00

0.17

0.35

0.48

0.10

0.14

0.07

0.14

2.45

Plumbing

1.00

.03

.22

.10

.05

.11

1.51

Heating

1.00

.10

.17

.07

.10

.06

.08

.02

1.60

Electrical

1.00

.18

.85

.13

.10

.13

2.52

Source: Housing Costs; Where the Housing Dollar Goes, p. 46.

In light of the potentialities for cost reduction offered by direct
purchasing, one might well ask why prefabricators have not been
generally more successful in cutting their finished materials and equip
ments costs. The problem involved here is not one peculiar to the pre
fabricators. It confronts the building industry as a whole and is a
crucial one in the evolution of large building enterprises of any sort,
whether of operative builders or of prefabricators. One conclusion
that has emerged out of almost every study of the building industry
in the last decade is that few if any of the industry's problems cannot
be traced in some way to the small scale on which operations are car
ried out; yet, when a building organization begins to grow to a
really significant size, it still cannot obtain many of the advantages
and efficiencies which ought to go with its stature. This has been
the protest of many of the big builders, of whom Levitt of Long
Island has been perhaps the most vocal, and it has also been the
protest of many a prefabricator. Even though prefabricators buy
materials in carload lots and even though they carry out most of the
functions which normally fall to the middlemen in the regular dis
tributive chain— principally stocking and maintaining a sizable in-

296

ventory, handling, sorting, grading— they still cannot, in most cases,
obtain mill prices. The reason is fairly obvious. Every time a pre-
fabricator buys at the mill, and at mill prices, at least one and per
haps two elements in the distribution channel (the wholesaler or job
ber and the retailer ) have been completely by-passed. The mills have
established stable business relations with their distributors, and they
are naturally not in a hurry to upset things by circumventing them;
the conventional building industry is based on the existence of these
long distribution lines. If prefabricators were producing 50% of the
housing in the United States instead of 5%, they would be so large
an element that the materials producers could hardly afford to ignore
them. But until the industry acquires that stature and until the firms
who produce more than 1,000 houses a year account for a major
portion of the annual building output, it is probable that the materials
producers will feel obliged to rely upon and support the conventional
distributors. The early phases in the development of big builders are
therefore likely to be the slowest.9 The importance of the large-scale
Lustron endeavor is great if from no other point of view than this.
As more big housing producers emerge, they will tend to become an
increasingly significant element in the industry, and the small builder,
to whom the present materials distribution scheme is well suited, may
tend to become increasingly unimportant. Neither trend was con
spicuous at the time of the survey.

III. Fabricated Components

In a manufacturing operation which consists largely of assembling
fabricated components purchased from specialized producers, the

9 The Architectural Forum feels that this trend is already established. In the
November 1947 issue, on page 10, it reports that its research, based on building
permits issued in 1946 and 1947, indicates that builders of 10 or more houses a
year presently account for three-fourths of all United States house construction,
whereas nine years before they accounted for less than half the houses built.
However, this statistic should not be interpreted without considering the fact that
housebuilding in 1946 and 1947 was proceeding at about twice its rate in 1938;
furthermore, there is a big gap between the builder of 10 houses a year and the
builder of 1,000.

297

distinction between procurement and production is very tenuous in
deed. This is, however, an important consideration, for not only
is there the question of how much of the house should be built in
a factory and how much at the site, but there is also the problem
of how much of the house should be produced in a central plant
and how much assembled from components of various manufacture
at one or more distribution points. The distinction is between the
prefabricator as a producer and the prefabricator as a synthesizer.
In either case he may retain the functions of design and integration
and of distribution. But there are a number of other considerations
which had led certain prefabricators to prefer one or the other of
these different types of operations.

By producing as much as possible of the house in his own plant,
the prefabricator will avoid paying for the overhead and profit of
other manufacturers; he probably will have greater freedom in de
sign; and he should have greater freedom in the administration and
control of his business operations. On the other hand, some of the
prefabricators visited had several reasons for freeing themselves of
as much actual production work as possible. For one thing, many
lacked the capital to build the necessary production facilities; this
was particularly likely to be true of a proponent of a metal house.
Furthermore, even if he had sufficient capital, the prefabricator might
well prefer to use it elsewhere and avoid tying himself to any par
ticular material or process. The period during which the survey was
conducted was one of great flux and high technological expectations.
It was natural that in such an atmosphere a number of companies
avoided heavy investment in plant and equipment and replaced
production with procurement so far as possible. Even in more stable
times highly centralized production might interfere with an objec
tive approach to design and deter the company from adopting new
materials or structural techniques simply because these would not
utilize existing facilities.

Perhaps the most obvious reasons for purchasing rather than pro
ducing certain components, however, were those of relative imme
diate costs. Component manufacturers might well achieve sub
stantially lower costs than a prefabricator making the same items
because of two advantages: specialized high-volume production on a
steady basis, and optimum location with respect to resources. It is
clear that some prefabricators did not have a large enough volume
to justify the purchase of a complex glue press, for instance. Fur
thermore, prefabricators had to locate these plants with reference
to their market as well as to such other factors as raw materials and

298

labor. The component manufacturer, on the other hand, worries
less about the location of the general housing market, locates near
the resources he uses, and thus avoids paying the transportation cost
for material that is lost in component manufacturing. Also, he main
tains a steadier production rate by selling to a much broader and
larger market. The lumber companies thus did a great deal of initial
processing for prefabricators. In the production of doors, windows,
and cabinets these advantages were great enough to warrant the
growth of a large industry specializing in this type of manufacture;
other examples of factory-made components which were purchased
by prefabricators included chimneys, stairs, and plumbing assemblies.

The decision to avoid heavy investment in specialized tools was a
fundamental consideration of the Harman Corporation in determin
ing its whole pattern of operations. Its plant in Wilmington, Del.,
served more as a warehouse than as a factory. There some fabrica
tion was done, such as of wood furring strips, but the operation was
generally one of storing and packaging the steel frames, panels,
windows, insulation, wallboard, and plumbing and heating equip
ment, all of which were being produced elsewhere by other firms.
In many respects the Harman operation was more like precutting
than prefabricating, but the example is illustrative of the assembly
operation. The HomeOla Corporation carried out a large share of
its manufacturing by subcontracting to firms near sources of supply.
A large Tacoma, Wash., lumber company assembled the modular
floor, wall, ceiling, and roof panels from its own lumber and plywood
and shipped these parts directly to the dealers. At its own plant
in Chicago, HomeOla manufactured some of the plumbing and equip
ment, assembled the heavier steel items produced by affiliated firms
in the area, and shipped this portion of the house package to the
local dealer at the same time that the wood portion was being shipped
from Tacoma. One Oregon prefabricator estimated that it would
entail about $75,000 in woodworking machinery and about $35,000
in assembly equipment to tool up for the production of stressed skin
panels, and largely on the basis of this estimate decided to subcon
tract his panel manufacture to one or more established manufacturers
in the region. On the opposite side, it can safely be said that the
tight capital position of Anchorage Homes, Inc., resulting from the
expenditure of an estimated one million dollars for a new factory,
was a major factor in its failure.

It is unnecessary to adduce further examples. As has already been
pointed out, even those companies which fabricated the entire shell
did not begin to manufacture everything in the complete house

299

package. There was no prefabricator who manufactured his own
water closet or heating unit, for instance. There were a few steps
in the direction of corporate integration, and there seemed to be an
interest in housing on the part of some of the largest steel companies
to parallel their general integrated expansion in the production of
consumer goods. At the time of the survey, however, no housing
analogue to the Ford Motor Company had appeared.

300

Part JL JL •

Chapter S

PRODUCTION

I. Plant Facilities

This chapter describes and analyzes the production operations of
the industry, taking into consideration the physical plant capacity, the
labor force employed, and other factors related to the amount of total
output. A description of some of the aspects of factory production
follows: the processes and equipment used, the plant layouts, the
scheduling, etc. Since the production aspects of particular kinds of
prefabrication have been well treated elsewhere,1 the chapter de
scribes only the general patterns of the industry's production opera
tions, discussing the reasons for some of the different patterns ob
served. A brief economic analysis explores such questions as the
quantity of manufacturing done by the prefabricator, the relationship
between costs and volume, the increase in productivity resulting from
prefabrication, and the cost structures typical of various groups in the
industry.

One sign of the stature of the industry is its growth in productive
capacity. In early 1948 this was estimated by PHMI as 120,000
houses per year, or more than three times the actual production.
For the bulk of the industry, however, it is known that estimates
of physical capacity at any time are not too significant because the
tooling-up costs are not high. The creation of capacity for the
early war housing and for the 30,000 houses which were to have been
produced under lend-lease in 1945 are examples of rapid expansion
under stimulus. It should also be noted that this estimate of 120,000
excluded many of those companies which had adopted unconven
tional approaches and were committed to highly mechanized opera
tions involving large investments in plant and equipment. In the
spring of 1948 the National Association of Housing Manufacturers,
representing these firms, estimated that if the potential capacity of
its membership were realized, it would exceed the then existing
capacity of the rest of the industry.

The value of plant and equipment of 40 member companies sur
veyed by PHMI in 1947 was $11,008,467, and the total assets of
these companies were more than $24,000,000. Although this was

1 Particularly in the Manual on Wood Construction for Prefabricated Houses.
See also N. S. Perkins, Construction Manual for Douglas Fir Plywood Dri-Bilt
Houses (Tacoma: Douglas Fir Plywood Association, 1940).

303

less than half of the $60,000,000 estimated total capitalization of the
industry, most of the balance was represented by the giant Lustron
Corporation. It is interesting to note, by way of comparison, that
by June 1948 Lustron alone had already contracted for some $12,-
000,000 worth of plant and equipment, an investment greater than
the 1947 total of the 40 firms mentioned above.

During the Bemis Foundation survey, the average size of plants
larger than 100,000 sq. ft., of which there were 16, was 223,000 sq.
ft. This does not include the Lustron plant, of which the floor area
was more than 1,000,000 sq. ft. The survey also revealed that 29
companies had completed new plant facilities since the war, and that
at least 22 more were planning or actually building new plants.

On the other hand, there were numerous companies, a great
many of which were not visited, which operated with quite primitive
equipment: a small shop or shed, some crude wooden jigs, and a few
power saws. Though they would have to be classified as prefabri-
cators, these firms were probably operating on a capital investment
about the same as that of a conventional builder with the same out
put.

There were 19 companies which had more than one plant facility.
In most of these cases, one or more of the plants was a materials
preparation or cutting organization near the source of supply. Sev
eral eastern firms had a lumber-producing and precutting plant in
the South and an assembly plant in the eastern seaboard area, for
example, Johnson Quality Homes, Inc.

II. Location of the Industry

The map (see Figure 34) shows the location of the plants of 82
firms reported to be in operation on January 1, 1948. It can be
seen that the industry was well represented throughout the eastern
half of the country and on the Pacific Coast. Relatively few firms
were found in the plains and mountain areas where the population
is widely scattered. The largest number of prefabricators were lo
cated on the Pacific Coast and in the Midwest.

Some of the reasons for this pattern of industrial location are
immediately obvious, but it may be worth while to explore the pos-

304

.£

sible effects of a number of factors: access to raw materials, access
to markets, access to labor supply, environmental hostility, and "ac
cident." Regarding the first of these, there is no overall pattern of
proximity to raw materials. It is true, as pointed out previously,
that there were 19 companies with two or more production facilities,
one of these being, in most cases, a raw materials processing point
near the source, but a very large number of firms were located at
considerable distances from their respective sources, whether these
were of lumber, plywood, steel, or aluminum. This may be explained
by the fact that the prefabricated operations did not, as a rule,
significantly reduce the weight of the materials going into his prod
uct (although his product was often significantly lighter than the
conventional) and, in most cases, did increase the total bulk. There
is an advantage, therefore, in being closer to the market than to raw
materials.

The weight and bulk of the house package make the transporta
tion problem such that the location of the plant relative to the market
was of primary concern. Although house packages have been shipped
as far as 1,000 miles and beyond, the vast majority were not trans
ported more than 300 miles for reasons of cost.2 We might thus
expect that prefabricators were serving local or regional markets
rather than national ones and that they were located close to where
houses were being erected; therefore they would be generally dis
tributed according to population over the country. In view of the
ease of entry into the industry and its fluidity at the time, we might
further expect that they were concentrated in areas where the build
ing activity was greatest, and to a large degree this was the pattern
observed.

If the concentration of plants is compared with the 1946 volume
of new private construction,3 the results show surprising agreement in
all but three regions. The concentration of prefabricators in the
Pacific Coast area and in the middle western states of Wisconsin,
Michigan, Illinois, Indiana, and Ohio appreciably exceeded the rela
tive volume of new private construction, whereas in the eastern states
of New York, New Jersey, and Pennsylvania the reverse was true.
Tending to favor prefabrication on the Pacific Coast were the strong
expectation of future markets based upon trends in population migra
tion, the rapid rate of growth of the metropolitan areas within which
substantially all the plants were located, and the predominantly

2 See Chapter 6, Marketing.

3 Construction and Construction Materials, Industry Report, Statistical Supple
ment, U. S. Department of Commerce (June 1948), p. 6.

306

single-family, open type of development which characterized the
area. The middle western states offered a mass market distributed
in urban concentrations of various sizes throughout the population
heart of the country, and a somewhat more receptive environment
than in most of the eastern states. Possibly the Middle West was
more receptive because some of the earliest ventures happened to
start there and have since proved themselves and demonstrated the
case for prefabrication; for those companies which came later, the
struggle with external obstacles was progressively easier and the num
ber of economic and other aids 4 progressively greater. In the eastern
states the expectation of future growth could not be compared with
that of the Pacific Coast, and furthermore, the population was largely
concentrated in a few highly developed metropolitan areas where a
smaller proportion of new private construction was in single-family
houses. General consumer resistance to the idea of prefabrication
in houses appeared to increase in the far eastern states, so that per
haps the greatest effort to conceal the prefabricated nature of their
houses was made by companies which had plants located in New
England.

The consideration of character of labor supply did not seem in
many cases to be an important one in fixing the plant location,
although several companies avoided highly unionized urban areas.
One reason for this is the relatively low proportion of the total house
package cost that was represented by direct labor cost. Another is
the fact that few special skills are needed in the average prefabrica
tion plant. Neither can plant location be explained exclusively as
the result of rational calculus. Personal preferences have been the
determining factor in more than one case.

Although there were some companies with two or more plants,
no company had a series of branch assembly plants. Serious interest,
however, was expressed by 11 manufacturers in the idea of branch
plants, not to carry on the full range of operations typical of a single
prefabrication plant, but to assemble components fabricated in one
or more main factories and left unassembled there for the sake of
economical shipping, and to act as warehouses for house parts com
prising a variety of designs. The principal difficulty was that an
investment in a chain of final assembly and warehouse facilities in
order to achieve wider and more economic distribution would require
more capital than any company had been willing, or than most had
been able, to risk thus far.

4 For example, a concern specializing in transporting prefabricated houses.

307

III. Labor Force

Statistics concerning the labor force must be interpreted with the
seasonal variations of the industry in mind. Even after prefabricators
have moved a considerable portion of the building process into a
factory, they are not completely independent of the weather. The
site has to be improved; foundations must be prepared; and some
time is required to shell in the house—almost always at least a day
and sometimes several days. As a consequence of factors such as
these, the prefabricators have been only partially successful in over
coming seasonal fluctuations, although there is reason to believe that
if dealers are trained and well enough capitalized to do more work
in advance, and if the proportion of site work decreases, these fluctua
tions will become smaller and smaller. It was the experience of
Gunnison Homes, for instance, that the active building season had
been extended by one and one-half months at each end, and it was
this firm's belief that it would be extended further. It should be
pointed out in this connection that Gunnison and a good many other
firms in the eastern and middle western states were attempting to
stabilize their factory operations by shipping house packages to
dealers in the southern states during winter months.

If one bears in mind the seasonal influence and the fact that many
of the plant visits were made during winter months, some idea of
the size of the labor force can be obtained. For the industry as a
whole, this was in the neighborhood of 10,000 at the time of the
survey.5 In those plants actually in production when visited, there
was an average employment of 79 workers. Similarly, the 1947
PHMI survey found a total of 2,810 factory workers employed by 40
companies in January 1947, an average of 70 workers per company;

5 The labor force at various times was reported as follows:

Number

Total

Date

of Firms

Employment

September 1946

115

10,200

November 1946

102

11,630

January 1947

167

10,450

Source: U. S. Department of Labor, U. S. Employment Service, Labor Market Infor
mation, Industry Series No. 24-33, Current Supplements for October and December 1946
and February 1947.

308

and the Department of Labor found an average employment figure
ranging from 63 to 114 at various times during 1946-1947.6 It is
interesting, and perhaps significant, that the average number of fac
tory employees per firm, as reported by PHMI, went from 70 in
January 1947 to 83 in July, to 98 in January 1948, and to 103 in
July 1948.7

Average capacity employment of 300 workers on a single-shift basis
was indicated by 24 companies; some of the estimates given may
have been overoptimistic. Only one company reported more than
1,000 as its capacity employment with present plant facilities. Again,
none of these figures included Lustron, whose projected output of
30,000-40,000 houses per year might call for a factory labor force of
4,000 spread over three shifts.

IV. Factory Processes and Equipment

This section describes the degree of industrialization found among
prefabricators and the factory techniques in use. Because the mate
rials used tend to be the most important factor in governing the
choice of the actual production techniques and tools, they serve
as the basis of organization of the discussion.

A. Wood

Wood is a material which has several advantages and a good
many disadvantages over other materials in its adaptability to indus
trial production. Perhaps its best quality is the ease with which it
can be cut, machined, and pieced together. Woodworking machinery
is inexpensive, at least by comparison with metalworking machinery.
It is therefore not necessary to reach such a high volume of produc
tion in order to put an investment in woodworking machinery on an
economic basis. A further advantage is that the production engineer
ing for most prefabricating in wood is relatively simple and does not

6 Loc. cit.

7 The 1947 figures are for 40 companies; the 1948 figures, for 50 companies,

309

require the collective effort of a staff of highly trained technicians.
On the other hand, compared to steel, wood offers a few distinct
handicaps to industrial production. It is, by nature, not so homo
geneous; it is dimensionally less stable; it is not so well suited to such
a process as the forced drying of a paint coat; and it cannot so easily
be shaped. Because of these qualities the production man dealing
with wood finds it more difficult to achieve good quality control, to
obtain close tolerances in dimensions, to benefit from the wide variety
of industrial finishes that have been developed, and to use high
speed material-forming equipment. For these and other reasons
there has long been a school of thought which holds that when real
housing industrialization does come, the basic material used will not
be wood.

1. Preparation and Handling of Materials

Many companies began the fabrication process with the manufac-
facture of lumber out of timber taken from their own tracts. There
were at least 38 companies which owned remanufacturing equipment
and were capable of creating finished lumber from large timbers.

Prefabricated house manufacture requires a fair degree of preci
sion in order to make the prefabricated components fit together
readily at the site. When wood is used as the basic material this
precision is not always easy to obtain, and it becomes important to
control moisture content in prefabricated house manufacture where
it might not be essential in conventional house construction. For
this reason at least 20 companies used their own dry kilns to bring
lumber to the desired moisture content before it entered the fabrica
tion process. A number of other companies used systematic air-
drying operations for the same purpose. The bowing out of plywood
panels is a possible consequence of changes in the moisture content
of the plywood after it has been glued into a panel, taking place when
interior and exterior plies of the panel expand or contract with re
spect to one another to produce a curvature. One way of counter
acting this tendency is to store plywood in such a way that its sur
faces are free to come to moisture equilibrium with a controlled
environment.8 Although plywood was almost always stored in an

8 It should be noted that, since the interior and exterior surfaces of a wall are
exposed to different environments, they will tend to attain different moisture con
tents in the course of time; consequently, if a stressed skin panel is built true at

310

inside heated space, only a few companies were known to be
following the practice of "sticking," which involves putting wood
strips between the sheets when piling them. Those that did had
less difficulty with panel bowing.

Only a few companies were found to be thoroughly inspecting
their plywood, and indeed there was no really good test for the
soundness of the glue lines. However, at least 13 companies were
dipping some or all of their framing members into a toxic preserva
tive in order to protect the wood from fungi or harmful insects. In
certain sections of the country, of course, this is more or less standard
practice.

The preparation of other materials was usually quite minor in
extent. Insulation and wallboards usually were purchased ready
to use, except for necessary cutting. Homasote, however, was wet
down to cause it to expand before being used for the surface of a
panel. It shrinks on drying, thus making the skin of the panel taut
and in effect prestressing it.

Only a few mechanized conveyer lines were seen in use to convey
lumber from storage to preparation points; this was being done by
forked lift trucks in at least 16 cases, by high-bodied carrier trucks
in at least six, and by gravity roller lines in two, but mostly by carts or
by hand.

2. Cutting and Machining

Prefabricates using wood usually required high-grade lumber and
great precision in their cutting and machining operations, particu
larly in the manufacture of plywood panels. In manufacturing these
the framing member had to be square with the plywood surface in
order to achieve a good glue bond. Additional precision was neces
sary at the perimeters of the panels where very accurate millwork
was often called for by the construction system. Furthermore, fram
ing members had to be quite straight, in order that they would fit
into the jig positioners properly. Thus, in a good many plants,
use was being made of large, high-speed, precision woodworking
machinery such as circular saws of various types, single and double

a time when the interior and exterior sheets of plywood have the same moisture
content, the panel may bow later on. The U. S. Forest Products Laboratory has
recently been conducting research to discover the optimum initial moisture con
tents for plywood sheets to be used in stressed skin construction.

311

planers, molders, and double-end tenoners. Other tools that were
sometimes found in use were multiple boring machines and multiple
dado machines. Molders, for instance, which are rather specialized
machines for the rapid production of framing or other stock cut to a
specified pattern, were known to be in use in at least 39 plants. The
design of plywood-cutting machinery showed great ingenuity, some
of the ideas having been developed during the war when production
often depended on the ability to improvise the required tools and
equipment. A few prefabricators were using automatic cutting ma
chinery designed and produced for that purpose by saw equipment
manufacturers, but the majority of those who were doing any extensive
plywood cutting had devised schemes of their own, using such ele
ments as traveling circular saws or moving tables. In many cases it
was desirable to give the plywood extremely accurate edge surfaces
and square dimensions, particularly when plywood sheets were used
to make up the inside wall and ceiling surfaces. To do this a double-
end tenoner was often used. This machine, which sizes sheets and
panels to precise dimensions, grooves panel edges for splines, cuts
stock to accurate length, and does many similar operations, is one of
the most versatile and one of the most expensive pieces of woodwork
ing machinery used by prefabricators in wood. At the time of the
survey double-end tenoners cost about $15,000-$20,000, and at least
14 companies were known to be using them. Another machine used
for accurate edging was the equalizer, and improvised machinery for
the same purpose was used in 12 cases. However, not all companies
using plywood as a surface material were sizing their panels or their
plywood in any way; some were content to rely on the accuracy of ply
wood mill fabrication to achieve reasonably good joints.

3. Subassembly

Where panels of room or wall size were being manufactured, it
was common practice to subassemble the framing members for
standard details such as door and window openings whose location
within the wall panel was not standardized. At least 19 of the
large companies utilized special jigs and tools to make up framing
subassemblies for standard openings, and thereby simplified assembly
operations when these were incorporated into large panels. In
fact, this manner of assembly usually made it possible to use fewer
jigs in the manufacture of a greater variety of wall panels.

312

This method was employed by Precision-Built Homes. All de
sign and production were based on the 4" Bemis module. Framing
subassemblies for windows, doors, and floor-wall plates were made
up on a set of standardized jigs. Job-lot orders based on almost
any modularized design could be rapidly manufactured in panels of
room size by the use of an adjustable "master" jig and the inter
changeable framing members and subassemblies which were already
in stock.

Several companies were carrying on an operation which might be
termed the subassembly of plywood. Accurately edged 4' X 8' ply
wood sheets were joined into room-size sheets by butting them over
a thin backing strip of plywood which presented a common gluing
surface for the contiguous sheets atop it. A strong glue line without
nails was achieved through the use of a fairly simple hot plate press
only a few inches wide which was operated on a fast cycle over one
joint at a time. The room-size sheets were then mounted on the
framing members of their panels with good assurance that joints
would not open. This made it possible to paper directly over the
joints without fear of cracks appearing later.

4. Assembly

Two principal means of fastening wood pieces together were being
used: glue and nails, frequently both. Hand nailing, of course, is
hardly an industrialized operation, even when done under a factory
roof, and it was most extensively used in those plants which pro
duced a panelized, but otherwise conventional, wood frame house.
There were, however, numerous attempts to simplify and speed up
the operation. A good many companies were using spring devices
which deliver a staple or nail into a sheet of plywood or wallboard
and into the framing underneath when struck on top with a sharp
hammer blow. A few factories used corrugated clips to assemble
their framing members, driving these in from above as the lumber
lay in the jig, instead of driving nails into the peripheral edges.
This made it possible to run the panel through some type of edging
machine without the danger of nails interfering with the process.
One pneumatic hammer was seen in use, and also one crate-nailing
machine, the latter for applying subflooring to 24' long panels. Sev
eral other companies were developing various sorts of automatic

313

nailing devices capable of either a sequence of operations or a num
ber of simultaneous ones.

On the other hand, a number of companies using nails to obtain
pressure for their gluing operations expressed the intention of in
stalling glue presses to replace nailing altogether. When nails are
used instead of a glue press, the bond is seldom so good, and conse
quently certain design advantages, such as the use of lighter framing,
may be lost. Other reasons for installing a press were the elimina
tion of nailing labor and material costs,9 and, if heat were used,
especially high-frequency induction, a decrease in the time required
for the glue to set. In order that these advantages be decisive, how
ever, they must outweigh the cost of an expensive piece of equip
ment. Only six companies were, in fact, known to be using hot
presses for gluing plywood to framing in the production of panels.
Heat was applied in four of these cases by high-frequency induction,
and in two by heated platens, one using steam for this purpose and
the other electricity. The largest hot press known to have been in
actual use was Prenco's, which had a 32' X 9' bed.

Being the newest of the developments, the high-frequency induc
tion technique was the one which had aroused most interest. Its
most frequently cited advantage over other types was the speed
and accuracy with which heat could be concentrated at the glue
lines. It could do this because the high-frequency electric field is
able to focus heat at a point well within the mass in which it is
oscillating, whereas the hot platen press depends on conduction of
heat inwards from the contact surfaces. One of the electronic presses
being used in the Midwest was bonding panels in about a minute,
whereas the steam-heated press required a three-minute cycle. Such
a comparison, however, is not particularly significant, because the
curing time depends very much on the glue used. Generally, with
either high-frequency or hot platen presses, the time required for
curing can be kept to a few minutes.10 A further device for speeding
up the gluing operation was the use of a multiple opening press,
such as the steam-heated one which was handling 10 panels every five
minutes, including loading and unloading. Such a press cost about
$35,000 in 1946, so that a fairly high production volume is necessary
if it is to be used economically. In this case planned production
was 16 houses per day. Other presses were designed for volumes of
40 houses per day, 35-40 per day, five per day, and in one case,

9 A not insignificant reason was the acute shortage of nails during the period of
the survey.

10 Manual on Wood Construction for Prefabricated Houses, p. 179.

314

only three per day. In the last two cases high-frequency machines
were employed, indicating possibly that for small, single opening
presses these gave greater initial or operating economy, or both, than
steam or electrically heated platen presses.11 Rapid curing was
also effected in a few cases by the use of heated chambers in which
glued assemblies were placed after clamping or nailing. Only one
company was known to be using a cold press in the production of
plywood panels. This technique, which saw some use during the
war, seems to have been too slow, and to have been generally aban
doned in favor of either hot-press or glue and nail techniques.

Most factories were using hand-operated glue guns to spread glue
onto framing members, although a few used glue-spreading ma
chines which applied the glue to both sides of a frame assembly
as it was fed between a pair of rollers. When panels had their
surface sheets bonded on simultaneously in a press, the use of such a
glue-spreading machine simplified operations considerably.

A distinguishing feature of almost every factory producing wood
panels was the use of jigs. Horizontal assembly jigs determine the
overall dimensions of panels without need of measurement, leveling,
or plumbing, and usually also determine the locations of members
or subassemblies within panels. The simplest jigs were crude wooden
tables utilizing rough wood blocks to position the framing members.
The more accurate and refined ones had steel tops and carefully
machined stops which provided for the easy entry of framing mem
bers, their precise alignment, and the quick removal of the as
sembled unit. Devices used to apply pressure to framing members
in order to hold them exactly in assembly position were stops acti
vated by compressed air, cams of various sorts, wedges, and screw
clamps. In designs where a high degree of precision was necessary,
as when both surfaces of a wall section were to be factory applied, it
was essential that the dimensions of the jig be very accurate and that
the members be squarely aligned. This could best be assured with
metal-based jigs, since the wooden ones had to be checked regu
larly for precision. At least 18 companies were known to be using
metal-based jigs. In some cases rather elaborate "master" jigs were
seen in use, these having a number of movable guides containing
notches or comb-like teeth to position the framing members within
the larger assembly. Such jigs were used where wall-size panels or
varied designs were being produced, as contrasted with the use of

11 It was possible to rent the electronic equipment and avoid a large cash out
lay. This may also have been a factor.

315

simpler jigs in the production of a rather limited number of types
of standard modular panels.

5. Finishing

The great majority of prefabricators in wood did not apply the
final finish to the main elements in the shell of the house in their
factories. Cabinets, trim, windows, and doors were usually sent
to the site already painted or stained, but the floor, wall, ceiling, and
roof were in most cases finished at the site. There were several
reasons for this practice. As was pointed out previously, some manu
facturers left most of the actual fabrication of the floor, ceiling, and
roof to be done at the site. Many companies planned to disguise the
panelized structure of their house by applying siding or shingles on
the exterior and by taping joints, hanging wallpaper, or even plaster
ing the interior. These operations were almost always done at the
site, although a few firms applied shingles and siding to panels in
the shop; therefore final finish coats were generally site applied.

In those cases where the wall or floor panels were completely
fabricated in the factory and where nothing but painting remained
to be done, it might seem at first glance most economical to do this
work in the shop. A big reason in favor of doing so is the large ele
ment of labor cost in a site-applied paint job— from two to five times
as much as the materials cost. However, there are at least two im
portant technical reasons for not doing the final finish job in the fac
tory. First is the danger of damage to the finished surface during
handling and transport and the expense of trying to prevent such
damage. (A partial solution to this problem would be to apply all
but the last coat in the factory.) The second reason is that a slow-
drying paint job means either a low production rate or else the use
of a large area in the factory in which to do the drying. These
problems might be avoided by the use of a fast-drying finish and a
method of forced drying. Unfortunately, however, there were not
available many fast-drying finishes suitable for exterior woodwork,
and a subsequent site application of such finishes for maintenance
purposes would be most difficult. In the forced drying of paint,
the shortcomings of wood for industrialized production are again
apparent: the high temperatures needed to speed the drying would
have to be limited by considerations of damage to the wood through

316

charring or excessive loss of moisture, and of damage to the paint
film by the expansion of air in the pores of the wood beneath the
coating.

Notwithstanding all these difficulties, final finish coats were being
applied to both exterior and interior surfaces in some factories.
In many more plants the woodwork received only a sealer or priming
coat, or both, on either one or both surfaces. And in a number of
other factories, no finishing of any kind was being done. Where
finishes were applied in the factory, the surfaces, usually of plywood,
were in many cases first machine sanded with drum or belt sanders,
then inspected and touched up where necessary, and sent to spray
booths. A few plants were equipped with automatic spray set-ups,
but most of those which did any spraying used manual equipment.
It was customary to use a conveyer line in conjunction with spraying
and drying operations. Other means of applying coatings at the time
of our survey included dipping, especially for sealers and water re
pellents, and brushes. Forced drying was sometimes done with
banks of infrared lamps; more often, by warm air.

6. Quality Control

One aspect of factory production which should not be overlooked
is inspection for quality control. Not only is this more readily done
in a factory than in the field, but it is also more essential to con
tinued business success for a prefabricator than for the average
conventional builder. Because of the infrequency with which houses
are bought, and because of the short-term interest of the average
builder and contractor in their product, the great bulk of home-
building has traditionally been carried on without the use of brand
names, quality guarantees, advertising, or servicing. Along with
the evolution of large operative builders and prefabricators there
has been a corresponding increase in the importance of establishing
a name and maintaining a reputation. Quality control is an essen
tial element in this process, and without it even the most extensive
advertising efforts may fail. Systematic attempts at some sort of
quality control were observed in almost every plant, and in a few
these were quite elaborate, ranging from the inspection of raw mate
rials through manufacturing inspection of dimensions, glue joints,
and machined surfaces to the final inspection of finishes.

317

B. Metal

Probably the best testimony to the admirable suitability of metal
for industrial production is its widespread industrial use. The rea
sons are not obscure: metal is abundant; it can be made homogeneous
to a high degree; its physical properties can be intentionally altered
over a wide range; it has good dimensional stability; it can easily
be formed by casting, forging, extruding, stamping, or bending; it
can be welded, soldered, brazed, riveted, or bolted together; and it
can be made to take a vast variety of finishes— vitreous enamel, paint,
lacquer, plating, and many others. Metal can be fabricated in a
great variety of ways at high speed and with excellent precision. Its
prime disadvantage from the standpoint of production engineering
is that metal-working machinery is generally expensive and often re
quires a skilled engineering force for its proper set-up, control, and
maintenance. Ordinarily plant fabrication of metal structures re
quires a larger plant investment than for wooden structures, and
production volumes must be accordingly higher before economies
are apparent.

The production of metal houses is described in less detail than that
of wood houses chiefly because there were many fewer firms pro
ducing them, and there was but little evidence of a general pattern of
factory operations. One other general remark is pertinent: most of
the metal house packages do not leave the factory in the form of
wall, roof, and floor panels as do wood houses. Only two firms
were known to be shipping fully assembled wall panels. Most of
the packages consisted essentially of separate frames and cladding,
the houses being primarily of frame assembly design.

1. Material Forming

Practically all the metal systems known to have been in production
utilized some sort of sheet steel or aluminum as the primary material,
both for framing members and for claddings. Even the most com
plete production systems began with the purchase of rolled sheet
metal. Flat sheets were punched and sheared as required, and then
sent into forming operations. The Byrne Organization and Harman
Corporation bought structural shapes already fabricated by other
companies; and load-bearing wall pans were being bought by Metal

318

Homes Company and The Steelcraft Manufacturing Company from
the American Rolling Mills Company.

In the prefabricators' plants, materials were made into structural
members, wall pans, and claddings of various sorts by the means
conventionally used for cold-forming sheet metal— roll corrugating,
press breaking, die rolling, die pressing, etc. Such operations gen
erally involve high tool costs and must, therefore, be undertaken
at a high production rate if unit costs are to be kept low. Lustron,
for instance, had installed about 100 pieces of press equipment which
cost roughly $3,000,000, but it predicated this investment on an an
nual production of as many as 40,000 houses. Since there are many
repetitive elements in the design of this house, some of these presses
were to be working at very high volume (see Figure 41).

2. Assembly

In addition to their forming processes, some plants carried on as
sembly operations, but these were usually of a minor and simple
nature. Fox Metal, for example, bolted together channels to form
I sections, and then attached connector angles to these. Stran-Steel
spot-welded sections together to form rib sections. General Homes,
which was one company that planned to make a complete wall
panel, bonded an aluminum skin to fiberboard sheathing and then
fastened this to a corrugated aluminum core. Lustron gang-welded
its structural shapes into roof trusses and panelized frames, and
assembled most of the elements for its bay window unit in the plant.
An exception to these essentially minor assembly operations was the
plan of Reliance Homes, Inc. (see Figure 42). Its scheme called
for assembly in the plant of complete house sections. Plant opera
tions included welding a basket frame of steel C channels, fastening
to it the interior surface of Homasote and exterior surface of alumi
num bonded to Homasote, and installing and finishing the floor and
mechanical equipment. Such a pattern is comparable in many ways
to that used in the production of the British AIROH house, and
together these two offer the best examples of the sectional house
worked out in metal.

319

3. Finish

Components of metal houses generally received at least a priming
coat in the factory, and often were completely finished there. With
steel, rust prevention is of course a major problem, and consequently
parts were primed with zinc chromate or some other paint as soon
as possible. The Harman house was of steel construction to which
only the zinc chromate priming coat was factory applied. Subse
quent coats of oil paint were sprayed on in the field, along with
mineral granules to give a stucco-like texture and somewhat improved
performance.

Complete factory finishing was more common than with wood
houses since the use of metals permitted accelerated drying and
baking of very hard finishes. The Lustron house was one of the best
examples of a completely factory-finished job, and it depended
heavily on the toughness of its porcelain enamel finish to prevent
damage to the surface during handling, transport, and erection. The
permanence of porcelain enamel is perhaps its greatest advantage, but
its application is confined to a factory where the necessary pulverizers,
dipping tanks, conveyer lines, and large gas or electrically heated
ovens can be located.

The Lustron porcelain enamel process was one of the important
influences in the development of the Lustron Corporation. During
the war, a method was developed for the low-temperature firing of
porcelain enamel. This was a "one-coat one-fire" process which
eliminated the base-coat operation. The process also permitted the
use of ordinary steel backing rather than the more expensive enamel
ing iron, since the lower temperature eliminated objectionable warp
ing that would result from conventional enameling of ordinary steel.

Structural members for the Lustron house were cold formed in the
plant from strip steel. These shapes were welded into wall panels
and roof trusses, and each assembly was given a protective coat of
enamel. The 2' by 2' exterior wall panels were stamped from light-
gauge cold-rolled strip steel, enameled, and insulated with Fiber glas.
All surfaces exposed to weathering action were given a special coat of
finish enamel in addition to the basic layer. The wall panels were
interlocking, and the joints between panels were sealed with a Koro-
seal gasket. In erection, the wall frame sections and roof trusses were
bolted together and then the cladding panels applied.

320

35 Ford house

36 Butler house

37 LeTourneau system

basic carrier of the Tournalayer
detail of inner form ready for pouring
outer form lowered over inner form
concrete pouring gun
pouring the concrete into forms
dropping the completed house at its site
finished LeTourneau houses

1 placing wall forms on floor slab

38 Ibec system

pouring concrete into wall forms

3 lifting wall forms from walls

***>*,

4 constructing roof slabs in stack

5 placing roof slab on Ibec house

6 completed house— Norfolk project

39 Gunnison plant operations

cutting plywood
cutting framing members

3 assembling panels for Gunnison house

4 trimming panels

5 bonding panels in multiple press

6 finishing Gunnison panels

40 National Homes plant operations

basic wall panel line, showing from front to
back: assembly of framing members in jig,
spreading glue on framing members, place
ment and stapling of interior wallboard, ap
plication of insulation, placement and nailing
of plywood exterior surface, insertion of win
dows and doors, and, finally, the completed
panel

2 floor panel line

3 roof panel line

4 gable-end panel line

5 special 16' double end tenoner

41 Lustron plant operations

1 fully automatic exterior wall panel press

* tf

2 grinding the frit

4 rolling Lustron frame members

3 mixing the enamel

5 baking enamel on Lustron roof panels

6 welding watt frame assemblies

7 loading wall panels on special trailer (note the extent of manual labor
involved)

8 loading frame assemblies (here plumbing watt panel) on special
trailer

"•TT

assembling

the

frame

applying

the

aluminum

surface

42 Reliance plant operations

finishing

the

complete

house

sections

43 Crawford Corporation— example of specialized woodworking machiner
used by large prefabricators for multiple cutting

44 Texas Housing Co.

Homette

standard house

C. Concrete

1. General Qualities for Production

The theoretical advantage of concrete from the production point of
view is one which exists for casting processes in general: ease of
forming to the desired shape. Because of the inherent simplicity of
casting, inventors have always hoped to develop materials or ma
chinery which, using this fundamental technique, would produce a
really economical house. If such a house is to be built out of pre
cast concrete units, however, the handicaps of long curing time,
frangibility, and weight must be overcome.

2. Preparation and Handling of Materials

The factory production of precast concrete units observed during
the survey was characterized by the extensive mechanization of proc
esses which were generally done by cruder methods in the field.
The bulky materials used were handled primarily by mechanical
batching equipment fed from rail-side hoppers or other types of load
ing machinery. Materials were mixed in various types of stationary
mixers. Since the use of a lower water/cement ratio results in a
higher strength, it was not uncommon for factory producers to mix
for longer periods than would usually be encountered in field prac
tice and to use less water, relying on mechanical equipment and
better-controlled factory conditions to vibrate thoroughly and to
handle a stiff mix with ease. Another preparatory operation handled
in the plant, and an important subassembly process in some instances,
was the cutting and assembling of the reinforcing mat. Further, the
use of special mix concretes, foaming or air-entraining agents, aggre
gates, or methods of mixing is ordinarily more feasible in the plant
than at the site.

3. Casting

From the mixer the concrete was poured into forms, which were
usually of steel. Some sort of vibrating table was generally placed

321

beneath the steel form in order, by vibrating the concrete, to achieve
more uniform densities and more precise dimensions. To put a
troweled surface on the top side of the casting, some type of machine
was used in almost every case.

4. Curing

In order to speed up the production cycle three methods for attain
ing early strength were seen in practice. Simplest, perhaps, was the
use of high early strength cement, a material which becomes strong
enough to withstand quite severe treatment in about one-fifth the
time required by ordinary Portland cement. A second method was
to pass the casting through a bath of steam under pressure. In the
manufacture of the Pfeifer concrete units, for example, slabs were
cured in a 36' long autoclave while still in their steel molds. A 12-
hour exposure to steam at 40 Ib. per sq. in. accelerated the curing
sufficiently to permit use of the molds on a one-day production cycle.

The third method, the Vacuum Concrete process, was based on the
utilization of atmospheric pressure and was significant not only be
cause of its high early strength, but also because it was helpful in
dealing with several other problems of precast concrete construction.
Until quite recently Vacuum Concrete had been used almost exclu
sively for heavy cast-in-place construction. At the time of the
survey, however, it had been used for housing of precast panel con
struction, primarily in projects of 100 or more dwellings. Casting was
usually carried out on or near the site, and although the necessary
equipment could be installed in a plant under cover, it was funda
mentally the same in all cases. The casting bed was of smooth con
crete, so equipped that air could be exhausted from a number of
grooves in its surface. When vertical side forms were placed over
these grooves and the suction turned on, atmospheric pressure held
the forms firmly in position. Hence they could be set and broken
away simply by opening or closing air valve connections to the large
vacuum pump that was the heart of the system. (Similarly, in the
field, joints between precast panels were formed with the aid of flat
or corner-shaped vacuum molds held in place by atmospheric pres
sure. ) After the forms were in place the concrete was poured into
the mold and was vibrated with a portable vibrator. Then one or
more vacuum mats were placed on the surface of the concrete, the
suction turned on, and water removed for 12-25 minutes. By lower-

322

ing the water/cement ratio in this way an unusually high early
strength was achieved and walls cast late in the afternoon could be
lifted into place the following morning. At the surface of the panel
the pressure of the vacuum mat and the extraction of much of the
water served to produce a particularly strong and dense concrete,
thereby increasing the resistance to moisture penetration. After the
mats were removed, the surface was troweled off to a smooth finish.

The same basic techniques were used for the floors, roofs, and
partitions. In some cases walls were cast in a sandwich manner:
2%" of aerated concrete would be poured first and allowed to con
solidate; an equal thickness of ordinary concrete would then be
poured and a slight amount of reinforcing embedded in this; finally,
the top surface would be troweled with either a pigment or a %"
layer of white cement grout to form an exterior finish. Such a 5" wall
had ample strength and good insulation properties.

The large slabs were lifted from the casting bed by means of
vacuum lifting mats which supported the weight of the slabs over
their entire surface and minimized any concentration of stress such as
would occur with ordinary sling lifting methods. The mats were
used in conjunction with crane equipment in much the same way
that an electromagnet is used to lift steel.

In precast systems in general, while it is true that special methods
were seen in use for breaking slabs out of their molds as soon as
possible after casting, it was nonetheless usually impossible to utilize
these slabs in construction right away; sufficient time had to be al
lowed for them to gain the strength required in handling and trans
portation and in carrying the designed loads in the structure. Hence,
a good deal of storage space in which the precast units could rest
while aging was generally required.

5. Tournalayer

Another type of concrete construction utilized the Tournalayer, a
huge machine developed in 1946 by R. G. LeTourneau. The Tourna
layer was used first as an outer form in pouring a monolithic concrete
house, and subsequently as a means of carrying the house to its final
site. If prefabrication is defined in its broadest sense as involving
the transfer to an off-site factory of a part of the construction process,
the Tournalayer falls under this classification, since the work of fabri-

323

eating and placing the forms has been almost completely moved away
from the site.

Because of the capital cost of the equipment involved and the ex
pense of transporting it over long distances, the Tournalayer has been
used only in large projects. In such projects a central operating site
was selected at the same time as the housing area. Here the steel
inner form, consisting of two chambered shells, was located. Pre
fabricated reinforcing steel, window and door bucks, electrical boxes
and conduits, and separators for wall endings were placed against
this inner form, and the four-sided outer form carried by the Tourna
layer was then lowered around the prepared core, usually leaving a
space of 5" for the casting of walls and roof. The assembly was then
ready for the concrete pour. If a high early strength concrete was
used, the house could be removed from the mold within 16 hours.
While still enclosed in the outside form the house was released from
the core by a lever mechanism which pulled in the sides of the core
about 2" all around. Outer form and house then were raised over the
top of the core by the Tournalayer and carried off. At the near-by
site, slightly excavated to receive the bottom edge of the wall slab
(which had been tapered outwards to serve as a foundation wall),
the Tournalayer lowered the house into its permanent position. The
outer form was then expanded, raised, and carried back to the casting
site by the Tournalayer, leaving the house ready for finishing details.
The typical house produced by the Tournalayer used 45 tons of con
crete and one ton of reinforced steel. Special concrete mixing and
pouring machinery was used at the casting site, and special cranes
assembled the interior forms. For obvious reasons, this equipment
has usually been rented, and not sold, by LeTourneau.

A number of projects have been built in the Southwest using the
methods described above. In such a climate, it is not necessary to
take many measures to improve the insulating properties of concrete,
and for this reason, as well as the very important one that the Tourna
layer requires a minimum of skilled on-site labor, it has been of con
siderable interest to builders in such parts of the world as South
America and the Middle East. It has aroused interest also because
of the surprising fact that it offers a good deal of diversity, being
able to make very different structures by simple rearrangement of the
forms, and even to cast two-story structures.

324

D. Honeycomb Core Sandwich Materials

Perhaps the most promising aspect of the various honeycomb core
sandwich materials 12 is that, through their use, prefabrication of the
house shell becomes very largely the production of one particular ma
terial. Designs embracing these new materials utilize the stressed
skin principle so extensively that there is a bare minimum of frame
assembly work to be done, either in factory or in field. There are still
such problems as window and door openings, and joints between
panels and at the floor and roof, but with the exception of a few sys
tems of concrete construction, most of which involved but little pre
fabrication, the development of the sandwich materials represented the
most direct attempt to change the building of the shell from a bits and
pieces assembly job to an automatic continuous material manufactur
ing process. Such processes had already been developed for certain
building products such as sheet metals and wallboards, but the manu
facture by similar methods of a composite material that would serve
at once as structure, insulation, enclosure, and finish still waits to be
achieved. The honeycomb core materials were not the only ones
using the sandwich principle. Cemesto, for instance, is a mass-pro
duced composite material consisting of a cane fiber insulation board
surfaced on both sides with a %" cement asbestos sheet, and combines
good insulation and surface qualities. It has been used only as a
curtain wall, however, not as a bearing wall. A related line of de
velopment has been pursued for a number of years by William B.
Stout.

Of the various types of core materials, plastic-impregnated paper
has thus far received the most attention.13 It is possible that such
paper cores may eventually be manufactured as separate materials to
which can be bonded surfacings of metal, plywood, paper-overlaid
veneer, or other types of laminates having the properties required
for stressed skin panels. At the time of the survey there had been no
mass production of these cores for use in housing, although processes
had been developed for similar materials in other uses, floors in air
craft, for instance. A good deal of development work had been

12 See "Physical Properties and Fabrication Details of Experimental Honey
comb-Core Sandwich House Panels," HHFA Technical Paper, no. 7 (February
1948).

13 Other materials which have been tried include plastic-impregnated fabrics,
foamed slag, foamed rubber, and glass.

525

done, however, and one factory was reported to be fully equipped for
the production of house panels of this material.14

Because it has not yet been worth while for the paper manu
facturers to produce a plastic-impregnated paper especially for the
purpose, the production of the core on a small scale, for development
work, began by treating a kraft paper with a phenolic resin solution.
The purpose of this treatment was to enable the paper to retain its
strength when exposed to moisture and to resist attack by decay or
fungi.

Fabrication of the core from the resin-impregnated paper then
could proceed by several different methods and could result in a
number of different types of core. One of the simplest processes be
gan by passing the paper, which was received from the mill in rolls,
through a corrugating machine and then through an oven to cure the
resin. After this the paper was cut into squares, or strips, and
passed through a glue spreader which applied glue to the nodes of the
corrugations. The sheets were then stacked in either of two ways:
with the flutes of adjacent sheets at right angles to one another or
with the flutes of adjacent sheets parallel and the nodes of adjacent
sheets in contact with each other, depending on the balance desired
between insulation and strength. The stacks were then put into a
press, after which they were sliced into portions of the proper thick
ness for a panel core.

Another method for making the core omitted the corrugating opera
tion. Sheets of plastic-impregnated and cured paper were striped
with glue lines spaced at about %" and were stacked and pressed
together with the glue lines of adjacent sheets parallel to each other
and staggered. After the glue had dried, the stack was simply ex
panded in the manner of a Christmas bell, and it was then ready to
have surface skins bonded to it. There had also been some develop
ment work on automatic core-making machinery, but none of the
methods in use at the time of the survey could produce in continu
ous strip.

14 The following are known to have been interested at one time or another in
honeycomb core sandwich materials:

( 1 ) Acorn Houses, Inc. ( 6 ) Kimberly-Clark Corp.

(2) Consolidated Water Power and (7) John D. Lincoln Furniture Co.
Paper Co. (8) St. Regis Paper Co.

(3) Chrysler Corp. (9) Southern California Homes, Inc.

(4) Consolidated Vultee Aircraft Cor- (10) United States Plywood Corp.
poration (11) Utley-Lincoln System, Inc.

(5) Forest Products Laboratory

326

The first method described was considered by many to be slightly
preferable from the viewpoint of production, because of its relative
simplicity and of the avoidance of any heavy investment in machinery.
It also afforded somewhat better insulative value.

Two types of equipment, roll and press, were used for bonding the
skin to the core, both of which were still being developed. The
Chrysler Cycleweld process utilized rolls and high-frequency induc
tion heating to produce a quick bond between core and skin, both
of which were preheated on their way to the machine. Should the
skin material come in continuous sheet form and should some method
be devised for making up cores continuously or as a continuous chain
of blocks, this process offered the possibility of the automatically
controlled manufacture of a standardized product— walls, floors, and
roofs "by the mile." On the other hand, the use of a large hot press
seemed somewhat more compatible with the production of panels
having various openings, edge fittings, and other specialized features
such as would be necessary at least in the production of walls. The
firm which had most closely approached commercial production of
houses made from this material was set up to use such a large hot
platen press.

V. Some Particular Aspects of Production

Prefabrication plants had several characteristics in common, re
gardless of the materials with which they were working, and these
are described briefly in the following section.

A . Factory Storage Facilities

One of the comments most frequently made by executives was that
if they had another plant to design they would certainly increase the
amount of space devoted to storage. At the time of the survey there
were at least two important factors contributing to the inadequacy of
storage facilities for raw materials. First, the building situation was

327

characterized by shortages, difficulties in procurement, and delays in
shipping. This often made it necessary for factories to carry unbal
anced inventories and to take materials as they could get them. Sec
ondly, production was seldom stable for a variety of reasons— inability
to obtain one or two items needed for the package, seasonal fluctua
tions which the distribution system had not been able to iron out,
and, most important, the failure of the marketing arrangements to
provide a steady flow of orders. In short, the materials-in-transit
concept of mass production had not been realized except by a very
few companies.

Storage space for manufactured goods was required for two prin
cipal reasons, to permit a relatively constant production volume with
a fluctuating rate of sales and to allow time for certain curing processes.
In the manufacture of plywood panels, for instance, recommended
practice was to allow a period of at least a few days in which the
glue could attain full strength and in which the water added to the
wood by the glue could be distributed uniformly. Precast concrete
units also required a curing period before they attained sufficient
strength for use. Of the 84 companies whose storage practices
could be ascertained, 40 stored panels according to type and made up
packages as orders came in; 24 stored their finished goods as house
packages; five stored them both ways, first according to type, then
according to house package; and 15 indicated that they kept no finished
inventory to speak of.

B. Plant Layout

Plant layout is an important aspect of production because it gives
an indication of the stage to which manufacturing methods have ad
vanced. Mass production involves two primary concepts: quantity
and standardization. The extent to which these concepts are realized
in prefabrication plants largely determines whether plant layout will
more closely approach line production or repetitive station production.

Line production may be defined as a method of manufacture or an
arrangement of work areas in which the material moves continuously
and at a uniform rate past a series of work stations and through a
sequence of balanced operations, thus progressing towards comple
tion along a reasonably direct path. In repetitive station production,
on the other hand, all the materials are brought to a number of work

328

stations at each of which one crew performs a complete sequence of
operations.

Thus, line production is characterized by a thin stream of material
which proceeds from the receiving department through fabrication,
assembly, and finishing to shipping along one line, or at most a few;
it involves a breakdown of operations into the simplest possible ele
ments and an extensive division of labor. Its advantages are many:
expensive high-volume production machinery can be effectively used;
there is a reduction in materials handling; more efficient utilization
of labor is made possible by greater specialization; supervision is fa
cilitated because delays are quickly detected and workers are paced
by the line; there is less congestion in the work areas; and the enforced
study of operations before the line is set up frequently results in in
creased efficiency. Against these advantages must be balanced a num
ber of limitations. A standardized output of reasonably large quantity
is required if labor and machinery are to be utilized economically; a
delay in the flow of materials to any point or in operations at that
point may force workers further down the line to remain idle; and
workers may be opposed to working on lines, especially if they are
accustomed to craft jobs; even if they are not opposed their productiv
ity and enjoyment of the work may be less.

For these and other reasons, many companies expressed a belief
in the economy of repetitive station production at the time of the
survey. In some cases their manufacturing process involved little
standardization of either house or components, with many types of
panels being produced, each at its own jig table. In other cases fluc
tuations in volume were severe, and more economical production
could be achieved by minimizing the investment in tools and allow
ing the labor force to vary with volume— the number of similar work
stations being increased or decreased as the occasion required. Occa
sionally the general scale on which production was carried out did
not warrant investment in conveyers, high-speed equipment, and
tools. These were some of the factors underlying the planning of
repetitive station layouts, several of them in plants producing at rela
tively high volumes : The Green Lumber Company, Hamill and Jones,
American Houses, and Pre-Bilt Homes Co., Inc.

It should be pointed out, however, that while these considerations
may have been applicable to some companies making use of wood
frame and plywood construction, the nature of the fabrication pro
cesses with other materials, notably metals, was such that it would
generally not be feasible to establish a repetitive station layout.

329

Some 53 out of the 103 prefabrication plants whose layouts could
be analyzed had more of the characteristics of line production than of
repetitive station. At least 37 companies were using conveyer lines,
and several were considering a type of layout in which jigs would
move past a series of work stations, rather than having the material
in process move down a line of jigs. Harnischfeger 1B was beginning
to use this scheme, and some of Lustron's assembly operations were
set up in a similar way (see Figure 45). Furthermore, many of the
companies which had turned to line production systems had previously
had several years' experience with less elaborate layouts more like
the repetitive station plan, for instance, Crawford, Green's Ready-
Built, Gunnison, Harnischfeger, National Homes, and Pease. Their
preference for the line production process may have been a sign of
some maturity in the industry and was certainly an indication that
many prefabricators had done a good bit more than just move a tradi
tional set of operations from the field into a factory.

C. Production Scheduling

Prefabrication factories were further distinguishable with respect
to their methods of scheduling production. On the one hand, there
were those plants which used what might be called a "job-lot" system
of timing, producing only after a definite order had been placed for a
specific job. On the other, there were those firms whose production
was scheduled on a more or less "steady-flow" basis, and which manu
factured standardized units somewhat in advance of specific orders,
maintaining a finished inventory of varying size. The job-lot schedul
ing system was widely utilized by those firms which produced a va
riety of designs, sometimes even individualized designs. For ex
ample, Precision-Built could take almost any floor plan, modularize it,
and produce it on what was essentially a line production set-up.
American Houses was also producing on a job-lot basis, but with an
assembly technique more like a repetitive station scheme, and with a
good deal of precutting and preassembly of standard parts. This
company's orders were usually for large projects, so that once a de
sign entered production, the firm could get some of the benefits of a
steady-flow basis. Another example of a compromise type of schedul-

15 An important reason for the Harnischfeger scheme was to increase the ca
pacity of a given size of plant.

331

ing was that offered by Better Living, Inc., which produced framing
subassemblies and floor, ceiling, and roof panels ahead of orders and
stocked them as standard parts, but which made up room-size wall
panels of varying dimensions only as actually ordered. Perhaps the
best examples of steady-flow scheduling were those plants which pro
duced modular panels somewhat ahead of sales so that an order for a
particular house could be filled from stock and shipped immediately,
as did Gunnison and Green's Ready-Built. At the time of the survey,
however, the overall demand was such and the materials shortages
were so great that these distinctions were often rather academic.

It is obvious that, for optimum production efficiency, materials
should be in continuous transit from receiving to shipping depart
ments, and that there should be an order outstanding for each piece
that comes off the line. This can probably best be realized if the
house is standardized, or if the components of a number of different
houses are all standardized. Not only does this facilitate efficient pro
duction, but the concept of interchangeability of packaged compon
ents is important also in shipping, for if a dealer is not ready to take
a house on a certain day because of weather, financing difficulties or
other delays, the same components, or almost all of them, can be used
somewhere else. Advantageous as it might be, however, steady-flow
production could not be carried on by pref abricators for more than a
limited period. Daily and weekly production economies were achieved
through steady-flow scheduling, but the leveling out of larger fluctu
ations arising from marketing difficulties, financing problems, and bad
weather could not easily be managed, and virtually no firm was in a
position to carry on full production in the face of seasonal changes
in building. To stabilize plant operations completely while being
subject to these and other disturbances in distribution would have
required far more capital than was available in most of the prefabri-
cation industry, and probably more capital than it would have been
economical to tie up for the advantages gained, even if it had been
available.

In making a broad comparison between the plants visited, it was
found, as might well have been expected, that there was some corre
lation between line production and steady-flow timing, and between
repetitive station production and job-lot timing. The correlation is
not so strong as one might expect, and yet it does serve to bring out
the fact that there are interrelationships among such factors as quan
tity, degree of standardization, extent of breakdown of operations,
division of labor, rates of material flow and of processing, and the
usefulness of specialized production equipment. It is because of the

332

interaction of all these factors that a prefabrication system becomes
as much a matter of industrial design as it is of architecture.

Number of

Layout Scheduling Companies

Line production Steady flow 38

Repetitive station Job lot 29

Repetitive station Steady flow 21

Line production Job lot 15

VI. Analysis

A. The Amount of Manufacture by the Prefabricator

An important fact in analyzing the contribution of the prefabricator
is the generally small proportion of the house which he actually pro
duces in his factory. This is a clue on the one hand to his inability
to achieve radical economies thus far, and on the other hand an
indication of some of the difficulties he faces in the field. There was,
indeed, wide variation among prefabricators in the extent to which
they carried prefabrication. This variation is a result of differences
in design, local building codes, local labor and building practices, the
size of projects, and other factors.16 Even so, a general statistic
will convey some useful information about the cost structure of the
industry. The average house package offered by 53 companies f.o.b.
factory represented 48% of the retail price of the erected house, ready
for occupancy, but exclusive of land cost. An analysis by the Office
of the Housing Expediter of cost breakdowns submitted by 12 appli
cants for guaranteed market contracts revealed a somewhat larger
percentage, 58%.17 Figures ranged from 37% to 77%, and in considera
tion of the inadequacies of data of this sort, it is necessary to generalize
that the average prefabricator was selling a package representing
roughly half the dollar value of the finished house, less lot.

16 Chapter 7 contains discussion of what is and what is not prefabricated under
various systems.

17 This cost breakdown is reproduced in Table 3. The information is also
summarized on p. 149 of High Cost of Housing.

333

Next, it may be asked what part of the value of the package is
"created by" the prefabricate! as a producer and what part is con
tributed by the materials used. Probably the most satisfactory meas
ure of this is the "value added by manufacture/' that is, the increase
in the total value of the commodities passing through the prefabri-
cator's plant as represented by the difference between the cost of the
materials consumed and the value of the products made from them.
The survey indicated that for the average prefabricator the value
added by manufacture amounted to about 35% of the house package
price. The figures of the Office of the Housing Expediter came to
roughly 40%. Again, in both cases, there was a wide range in the
data, from 25% to 45%, due largely to differences in design and in the
relative amounts of jobbed materials and processed materials going
into the package. As the package was composed more of materials
which the prefabricator simply bought, stored, and packaged, and less
of materials which he actually processed in his plant, the prefabri
cator became a synthesizer and distributor rather than a producer,
and the lower was the percentage of value added by manufacture.

Noting that the house package represented about 50% of the value
of the finished house and that only about 35% of the package was
"created by" the prefabricator, we can deduce that his contribution,
as measured by the value added in manufacture, is only about 18%
of the retail price of the house. When we compare this figure with
the percentage of value added by manufacture in several other indus
tries, we see that it is quite small: automobiles, 32%; furniture, 49%;
lumber and basic timber products, 56%; machine tools, 70%.18 This
puts the prefabricator in a difficult position for, supposing that by
some means he is able to cut his production costs in half— no mean
feat— he will have reduced the cost of the finished house by only 10%
(setting the percentage of value added at an even 20%). In fact, his
contribution is so small that his production position, from a cost point
of view, might be termed precarious. Although this situation may not
have been too well understood by some of the more enthusiastic pro
ponents of prefabrication during recent years, it was pretty generally
appreciated by members of the industry.

Prefabricators are attempting to do a job in the factory that has tradition
ally been done in the field. When that job is moved from the field to the
factory, overhead zooms upward. The small builder has practically no

18 Census of Manufacturers, 1939, Vol. II, Part 1, pp. 509, 549; Part 2, pp.
431, 522. Values cover the period from 1931-1939. The comparison is some
what unfair because these figures represent per cent of wholesale price.

334

overhead by comparison. In the factory we have, or should have, low
labor rates, but overhead compared to that in the field is multiplied many
times. That means we must do the same job more cheaply in the factory
after accounting for a greatly increased overhead. Of course, we should
be able to furnish materials more cheaply than the builder can buy them
himself, but it is also desirable that we do the work on the materials more
cheaply than the builder can do it. Put it this way: the value added to
the materials by the manufacturer is a small part of the total value of a
house. Since it is a small part of the total, the savings on this segment
of building cost must be decisive and must be demonstrated.19

The desire to increase their contribution to the total value of the
house, and to achieve the potential economies therefrom, was a major
factor in causing many prefabricators to manufacture items which
were subsidiary to the main structure, doors, cabinets, closets, etc.
There were other reasons, too: to assure a steady supply in a period
of shortages; to obtain the exact dimensions and specifications neces
sary for a certain design; and to utilize scrap pieces, for instance by
gluing them together to make counter tops for kitchen cabinets. There
were, on the other hand, manufacturers who maintained that it was
more economical to buy such items from specialty houses, or that it
would soon be more economical to do so because of the production
efficiency that went with such specialization. In any case, the follow
ing numbers of companies were found to be manufacturing various
subsidiary items:

Number of

Item Companies

Window sash 38

Trim 37

Doors 36

Kitchen cabinets 31

Plumbing assemblies 27

Wardrobes, closets, or storagewalls 17

Sheet-metal ductwork 13

Flooring (softwood) 3

Whether through production or purchase there seemed to be a
trend towards supplying a more and more completely prefabricated
house. This was not a trend which could be positively ascertained,
since the survey was, by its nature, a look at the industry at a definite
time. Yet the expressed expectations and intentions of many prefabri-

19 From a talk entitled "Uniform Cost Accounting for Prefabricators," by Wil
liam A. Tucker, Statistician, PHMI, at 5th Annual PHMI Meeting, March 31,
1948. This statement points up production problems, but perhaps minimizes the
savings possible in overall integration of the housebuilding process, from pro
curement to final financing and erection.

cators lay in this direction. The reasons given related primarily to
the economies which they hoped to realize through greater efficiency
in procuring various components and materials and in production,
through less site wastage, fewer setbacks due to weather, and less
time lost by having one crew wait for another to finish its work.

It was not demonstrated, nevertheless, that the greater the degree
of prefabrication, the lower the costs; the optimum degree of pre-
fabrication was not established. Certain designs were largely de
pendent on the use of factory processes, while others were quite as
easily fabricated in the field as in the shop. In the latter case, a
question such as whether roofs should be panelized or precut, or per
haps not furnished at all, was one which hinged to a great extent on
conditions at the site: the cost of field labor relative to factory labor,
the skill of crews available for erecting the house in the field, the
conditions of weather and land at the site, the cost of supplementary
materials in the field, the transportation costs from plant to site, the
number of houses being erected in one group, etc.

For wood frame designs there seemed to be an inverse relationship
between the amount of fabrication economically performed in the
plant and the number of houses to be erected in one group. Ameri
can Houses, for instance, was fabricating oply about 40-45% of its
structure in the factory, but seldom sold house packages for erection
in groups of less than 100. The Byrne Organization's 1,200-unit
Harundale project (not a wood frame design) utilized careful cost-
accounting methods to divide the work between site and shop;
roughly one-fifth of the total man-hours per house were performed
in the shop, the balance at the site.20 The abandonment by Kaiser
Community Homes of the prefabrication of wall panels in favor of
precutting 21 is a further indication of this point. Other large opera
tive builders such as Levitt, Bohannon, and Ponty seemed to find that
some combination of precutting the main structure and prefabricating
minor components gave the most economical results, and much of the
war experience with large projects pointed to similar conclusions. The
major reason was, of course, that in such projects many aspects of
mass production could be achieved without entailing the overhead
and distribution expense that burdens the prefabricator. There could
be mass purchasing, use of jigs and high-speed cutting equipment,
and an extensive division of labor among crews that move from

20 High Cost of Housing, p. 168. Actual figures given, 207 man-hours in the
shop and 797 at the site, were proved low by later accounting. Final figures were
not available.

21 This change in pattern of operations occurred after the survey was completed.

336

house to house, rather than having the houses move past them on an
assembly line.

On the other hand, single-house or small group erections did not
offer these opportunities, and it was argued that for such projects
much more fabrication should be done in an off-site factory where
mass-production techniques could be used. Whereas one firm which
precut and erected its own house reported that it cost $1,200 less per
house to build in groups of 10 or more than to build a single house,
the cost differential which could be obtained from erecting a highly
prefabricated house, such as a sectional type, in groups rather than
singly was probably quite small. Certainly, in such a case as the TVA
sectional house, where shipping and field assembly accounted for
only 12% of the total costs, or the British AIROH house, where these
items were estimated to comprise only 9% of the total, the economies
of large projects could not be too important.22 This argument cannot
be carried too far, however, for the costs of grading, installing utilities,
and constructing the foundation could be appreciably lowered in large
projects. And if, as in some sectional house systems, heavy equipment
such as a boom crane was required at the site, further economies could
be realized through large group projects. Finally, it goes without
saying that the larger developments improved lots more cheaply. Not
withstanding these qualifications, it seemed a reasonable hypothesis
in general that the larger the number of houses to be erected in one
group, the less the optimum degree of prefabrication.

Probably as important a factor in governing the amount of prefabri
cation as any of the above was the existence of many problems of a
"political" rather than a technical nature, including such practices as
local purchasing to appease local distributors, and the elimination of
certain items because of the wide diversity in codes. The solution of
these problems will require much time and effort; undoubtedly the
attention which they have recently received has been helpful. When
consideration is given to this factor and to the steady, if slow, prog
ress in materials and structure through research and development,
there was evidence of a trend towards more complete prefabrication,
at least of major components. This trend seemed most noticeable,
and most logical, where the newest materials and structural systems
were involved.

22 "Total cost" here excludes cost of land, grading, utilities, and foundation.
The last three of these items for the average AIROH house totaled twice as much
as the shipping and site assembly costs. For TVA cost breakdown, see Table 5.
For AIROH figures, see Table 6.

B. Production Volume and Production Costs

Before turning to an examination of actual production costs it
would be useful to know how costs varied with volume and to what
degree prefabricators were successful in achieving one of the pre
requisites for mass-production economy: high volume.

The volume at which major production economies began to be
possible was not easy to specify. It depended largely on the nature
of the house, the materials of which it was made, and the extent to
which it was composed of repetitive elements. Thus, one manufac
turer of a panelized but otherwise quite conventional wood frame
design reported that he would make no profit if he produced one house
per day, $18,000 per month if he produced two houses per day, and
$45,000 per month if he produced three houses per day. On the
other hand, one of the manufacturers of stressed skin panel houses
was operating with a break-even point of four to five houses per day.
The break-even point for a venture such as Lustron was probably be
tween 30 and 50 houses a day, compared with its capacity of 100
houses per day, on which figure its pattern of operations was predi
cated. Furthermore, it was difficult to untangle such factors as the
importance of other manufacturing operations where prefabrication
was only a subsidiary one. It was clear, for instance, that a lumber
and millwork company which carried on a subsidiary prefabricating
operation would have a different cost picture from that of a company
whose sole work was prefabrication. The former might achieve
economies through bulk purchasing of raw materials and through in
tensive utilization of production equipment simply because of the
large-scale manufacture of millwork, and not at all because of its
prefabrication volume, which might be quite insignificant by com
parison.

In the light of such wide variations there was no single volume at
which mass-production economies began. It is a fundamental char
acteristic of industrial production in general that as volume increases,
up to a point, unit costs decrease. No company reported that it was
operating in the range where increasing outputs would no longer
yield decreasing unit costs. The question might better be put, what
volume was necessary to attain an important share of the economies
deriving from mass production?

Houston Ready-Cut felt that it did not begin to achieve maximum
economies at less than 2,500 units per year. W. W. Rausch, then of

338

Anchorage Homes, said that he believed an annual production of at
least 10,000 houses per year was necessary for full production econo
mies with wood. C. W. Farrier, a former Technical Director of NHA
who served as Housing Research Director for Gunnison, reported
that "some of the prefabricators whom I have talked to indicate that
the volume of houses that they will have to turn out in the plant in
order to have sufficient ordering power to get reduced prices on
materials amounts to somewhere between 20 and 25 houses a day" 28
(5,000-6,250 units a year). A British writer, D. Dex Harrison, said,
"It seems likely that the specialized designs will require a minimum
of 5-10,000 [units per year] before the economy of mass production
is achieved in the house as a whole and before variations on the one
design can be contemplated." 24

These estimates, all but one of which are explicitly based on wood
as a material, average about 5,000. No company, in either 1946 or
1947, produced this many houses; the largest annual volumes reported
were between 1,500 and 2,500.25

As for the industry as a whole, it was operating at somewhat less
than half its estimated capacity in 1946 and 1947. Between October
1946 and June 1947, 87 plants were visited which were actually in
production. Of these, 27 gave no estimate of capacity, and the re
maining 60 reported that they were producing at an average of 38%
of stated capacity.26 This, of course, must be evaluated with the sea
sonal pattern of building in mind. A winter slump is customary,
even in prefabrication. In 1946 the industry produced 37,200 houses,
and in 1947, 37,400.27 At an average package price of $3,500, this
would mean a gross dollar volume of about $130,000,000. The 1947
total, however, represents the output of considerably fewer firms, so

23 Proceedings, American-Soviet Building Conference, p. 50.

24 Harrison, "An Outline of Prefabrication/' in Tomorrow's Houses, p. 132.

25 1947: American Houses, 1,600; National Homes, 2,500; Kaiser Community
Homes, 2,500.

26 Breakdown of 60 companies according to reported capacity:

Over 200 houses per month 5

100-199 9

50-99 13

25-49 11

Less than 25 22

27 1946 total by the Office of the Housing Expediter, 1947 total by PHMI,
PHMI News Release, May 3, 1948.

339

that average production per firm rose substantially.28 Even so, at the
beginning of 1948 PHMI estimated the industry's existing capacity
at 120,000 houses per year,29 so that, based on this figure, the produc
tion for the previous year was at less than one-third of capacity.30
It is therefore safe to generalize that the industry as a whole was not
utilizing its plant facilities to the optimum extent and that there were
potential production economies which had not been achieved.

C. Productivity

One way of measuring the potential effectiveness of prefabrication
in reducing costs is to find out how much it increases the productivity
of labor or, put another way, decreases the number of man-hours re
quired to build a house. For this purpose two statistics can be used:
the number of man-hours of direct factory labor per house, and the
number of man-hours of direct site labor per house. These should be
qualified, however, by differences in the size and quality of the house,
by differences in what is included in the prefabricated package, and
by differences in the amount of the package which the prefabricator
procures rather than produces himself.

For 29 producers of wood houses there were required an average
of 226 man-hours in the factory. Figures ranged from 100 to over 600.
Other studies of prefabricators working in wood have yielded results
of the same order of magnitude. The 1947 PHMI survey of 40 mem
ber companies gave an average of 268 factory man-hours per house,
and a study by the Bureau of Labor Statistics 31 of 14 prefabricated
war housing projects found an average of 242.

At the site, 26 companies were found to require an average of 238
man-hours to erect and complete the house, exclusive of subcontracted
work such as grading, foundation, heating, wiring, plumbing, and
sheet-metal work. PHMI found in its 1947 survey that an average of
182 man-hours was consumed in erecting the house and an average

28 OHE figures for 1946 include shipments by 198 producers. The 1947 figures
are based on shipments of approximately 80 companies.

29 Statement of Harry H. Steidle, Manager, PHMI, before the Joint Committee
of Congress on Housing, January 14, 1948.

ao Actual production in 1948 was 30,000 units (PHMI News Release, June 4,
1949).

31 Alexander C. Findlay, "Construction of Prefabricated and Conventional War
Housing Projects," Monthly Labor Review, 63 (November 1946), 723, 727. See
below, pp. 342-4.

340

of 276 in finishing it, making a total site time, exclusive of work on
the lot and foundation, of 458 man-hours per house. The BLS study,
based on large projects where many economies at the site were pos
sible, found an average of 440 site man-hours per house, but this
figure included work on grading, utilities, and foundation. All the
above figures were for wood houses of various designs, except that
no sectional types were included. There were some well-publicized
demonstrations in which a house was erected in less than half a day
by a few men, but when these were examined more closely it could
be seen that a good bit of preparatory work had been done in im
proving the site, building the foundation, and, often, in having special
pieces of equipment ready to do their special jobs. With the sectional
house, the time required for erection was at a minimum. The Reli
ance house, for instance, was completely erected in a demonstration
during a snowstorm in less than 20 man-hours.32 The Prenco and
TVA houses generally were erected, complete with all connections
made, in one day or less, using a crew of six to eight men. The
AIROH house, a British sectional type of which more than 69,000
were built since the war, required less than 50 man-hours 3S to erect,
whereas other prefabricated houses built under the British Temporary
Housing Program which arrived at the site as collections of panels,
cabinets, subassemblies, and loose material required an average of
300-400 man-hours 34 of site labor. Again, these figures do not include
the work of preparing the site and foundation and of installing utili
ties. A cost analysis of the AIROH house, for instance, shows that
these three items may total over four times as much in cost as the
erection itself.35

If allowance is made for these differences in basis of figuring among
the companies, then for the "typical" 24' X 32' house of panelized
wood construction roughly 250 man-hours were required at the fac
tory and 450 at the site (not including grading, utilities, and founda
tion). It would be interesting to compare these figures with com
parable figures for convential construction, but it is difficult to obtain
productivity data for conventional building which would permit a
fair comparative analysis. Not only should such data be classified
according to the size and quality of the house, but also according to
the number of houses built in any one project, the conditions of

32 Near Philadelphia, Winter 1948.

83 Unpublished paper by Carroll A. Towne, Prefabrication Advisor, HHFA,
May 1948, in the files of the Bemis Foundation.

84 Loc. cit.
35 Loc. cit.

341

weather, materials supply, and so forth. Estimates range from 1,000
to 2,500 man-hours as the labor time required in building one house
by conventional methods, but the exact basis of these estimates is not
clear. In the absence of results of controlled experiments, it will
have to suffice to use what would seem a reliable figure and a fair
one for purposes of comparison: the estimate developed by the Small
Homes Council of the University of Illinois in its time-study analysis
of the construction of the "industry-engineered house."36 This was
a two-bedroom, single-story, 768 sq. ft. dwelling with basement. The
total requirement by conventional methods averaged 2,091 man-
hours,37 and, according to the report, "Records indicate that savings up
to 20% of total labor can be made by the use of engineered con
struction methods and organized operations at the site." 38 Figures
included all work from excavation for the basement to finishing de
tails. They were, furthermore, based on the construction of one
house at a time.

When housing is built in large group projects, however, productiv
ity comparisons are apt to yield quite different results. Probably the
best study of this sort was one made by the Bureau of Labor Statistics
based on 24 war housing projects, two-thirds of which were prefabri
cated.89 It was found that the average saving in total man-hours at
the prefabricated projects was only about 8% (p. 343). All the projects
studied used wood as their basic material, but the prefabricated group
was further classified into three different types: stressed skin, frame
panel, and incomplete prefabrication (the last subgroup included two
frame assembly and two frame panel with conventional floors and
roofs). Man-hour requirements for these three classifications were
found to differ significantly: for the first they were nearly one-quarter
less than for conventional construction, for the second about 2% less,
and for the third about one-sixth more. The comparisons were for
corresponding operations— the customary site work at the conven
tional projects and the site work, plus factory work, plus related oper
ations such as transportation at the prefabricated projects. As the
study was careful to point out, however, the data used were insuffi
cient for general comparisons between prefabricated and conventional
construction. For one thing there were differences in weather, in

36 Research Report on Construction Methods.

87 Of this total, excavation, footings, foundation, basement, floor, floor joist and
subfloor accounted for 257 man-hours.

38 Research Report on Construction Methods, p. 16.

39 Findlay, op. tit., pp. 721-32.

342

the "natural" efficiency of labor,40 and in materials supply conditions
in different regions. Furthermore, the data applied to housing built
in large group projects so that, in effect, the word "conventional" had
a rather special meaning. Lastly, it would be unfair to judge present
prefabrication by the wartime product.

Unit Man-hour Requirements on War Housing Projects (by type of
construction)

Man-hours

Requirements as a Percentage of
Conventional Requirements

i ype or
Construction

Per Dwell
ing Unit

Per 1,000
Gross
Sq. Ft.

Per Dwell
ing Unit

Per 1,000
Gross
Sq. Ft.

Composite

Conventional

917.2

998.8

100.0

Prefabricated

682.1

978.3

74.4

97.9

92.7

Stressed skin

564.3

798.7

61.5

80.0

75.9

Standard panel *

716.3

1,041.7

78.1

104.3

98.5

Incomplete 2

877.0

1,236.0

95.6

123.7

117.3

1 Frame panel according to our classification.

2 Includes two frame assembly and two frame panel with conventional floor and roof.
Source: Alexander C. Findlay, "Construction of Prefabricated and Conventional War

Housing Projects," Monthly Labor Review, 63 (November 1946), 721-32.

In spite of all these variables, however, mention should be made of
some of the conclusions of this study:

(1) Of the total man-hours required for site work, from a fifth to
three-fifths could be transferred from the site, to be carried on in
the prefabricating plants and in related operations such as transpor
tation. In no case, however, was the site work reduced to a negligible
figure, the lowest being 261 man-hours.

(2) As for skills, the prefabricated projects required a larger per
centage of laborers' and foremen's man-hours than conventional oper
ations. However, it was found that the prefabricated and conven
tional operations were not so dissimilar in percentage distribution of
skilled workers and foremen by trades.

40 For instance, differences in standards and in the "natural" efficiency of labor
made the data biased in favor of stressed skin and against incompletely prefabri
cated houses.

343

(3) The comparatively few instances in which non-structural work,
such as plumbing and electrical work, was performed in the prefabri
cating plants demonstrated that a net man-hour saving could be ex
pected from such plant operations only under certain circumstances:
when there were a minimum number of connections to be made
between panels; when work could be concentrated in a small portion
of the house, for instance within one or two adjoining panels; and
when excessive protection or care was not required to prevent damage
during transport.

D. Production Costs

Turning now to a consideration of costs, probably the single cri
terion by which prefabrication has most often been judged, the final
production cost of a prefabricated house should, ideally, be compared
with that of a conventionally built house of the same size and quality
in the same location. Further, a determination by accurate account
ing of the optimum degree of prefabrication, qualified according to
the type of market, the design, the number of houses being built in
one group, and many other factors, would be desirable. But the main
interest of the consumer is the price of the house, and price involves
many factors in addition to production costs factors such as market
size and location, dealer organization, transportation, and financing,
which are covered in the next chapter.

Unfortunately, reliable cost breakdowns are difficult to develop.
For one thing, manufacturers were understandably reluctant to re
lease the information. For another, cost-accounting systems were not
uniform, the same item being counted in a number of ways by differ
ent prefabricators. (Recently there had been an effort, led by PHMI,
to standardize cost-accounting practices so that prefabricators could
compare cost figures and learn from each other.) And third, com
panies varied greatly in the extent to which they acted as jobbers, in
the amount of production which they did themselves, and in the com
pleteness of their package. For these and other reasons outlined be
low, such cost breakdowns as can be presented in a publication of
this sort are of somewhat limited value.

344

1. Cost Figures Submitted to the Office of the Housing Expediter

Tables 2 and 3 summarize data submitted to the Office of the
Housing Expediter during late 1946 and the first half of 1947. This
information should be interpreted with the following facts in mind:
the sample was a very small one; the figures do not reflect today's
prices; and the data were really estimates of cost made by firms which,
for the most part, had done little or no previous work in prefabrica-
tion— therefore some of the figures might better be regarded as decla
rations of intent than as records of performance. A digest of these
figures appears in High Cost of Housing along with a commentary
written chiefly by the staff of the Housing and Home Finance Agency.
We quote:

The direct factory labor costs range from a low of 1.13 percent to a
high of 14.10 percent. The low percentage is found in a plant which sub
contracts virtually all of its fabrication, and therefore a fair median per
centage would be closer to 12 percent than the average of 7.48 percent
shown in table 2.

... It is obvious that prospects of securing cost reductions through
elimination of direct and indirect labor in plants are definitely limited. For
example, cutting the direct factory labor cost in half would reduce the
total cost of the erected house by 3 to 6 percent. Reduction of field labor
costs, which range from a low at 7.81 percent to a high of 26.17 percent,
perhaps holds more promise.

On the other hand, the direct materials cost in the house package is, in
every case, the highest single factory cost item. In most cases, this is
true in the field as well. The prospect of savings here, both by develop
ment of designs which eliminate unnecessary material, and by reductions
in unit materials prices through elimination of wholesale mark-ups are
substantial.

Indirect and administrative costs generally represent such small per
centages as to offer little promise of cost reduction. It should be noted
that allowances for factory sales expense are abnormally low in every case.
Informed judgment on this subject has concluded that a factory sales
allowance of 5 percent is essential to successful merchandising in this field.

Considered either separately or combined, the factory and field allow
ances for profit in this tabulation cannot be regarded as excessive. . . .
Actually, the average field profit of 8.84 percent is somewhat below that
which is customary in the field of conventional building.41

It may be interesting to note, by way of comparison, that the evi
dence presented to the Joint Committee on Housing of the 80th Con
gress 42 indicated that, for conventional residential construction, labor

41 High Cost of Housing, p. 151.

42 High Cost of Housing, pp. 76-9. Sources quoted: Housing and Home
Finance Agency, Bureau of Labor Statistics, New York City Housing Authority,
The Econometric Institute, Inc., New York. Much of the testimony is conflicting.

345

Table 2

Summary of Unit Costs of Prefabrication Price Ranges

Total

Plywood

Element of Total Cost

From

Average

From

Average

Package, total

2,269.13

5,794.57

3,460.67

2,747.26

5,794.57

3,799.57

Erection, total

1,150.00

3,824.00

2,448.27

1,150.00

2,657.83

1,828.99

Total cost, less land

4,468.19

7,702.42

5,908.94

4,468.19

7,518.57

5,628.56

Package

Direct material

1,243.98

3,081.00

2,110.37

1,785.07

3,081.00

2,344.41

Direct labor

63.60

1,060.29

451.08

163.44

1,060.29

700.48

Indirect labor

65.36

604.39

183.82

111.63

185.38

143.27

Other indirect

116.30

577.18

331.42

284.30

577.18

431.01

Administration

29.94

197.00

97.30

29.94

52.28

39.88

Sales expense

37.57

258.98

151.38

152.40

258.98

200.41

Profit

226.91

579.46

354.70

256.03

579.46

399.11

Erection

Direct material

232.00

1,905.00

841.85

232.00

831.99

401.44

Direct labor

422.00

1,471.44

820.35

422.00

986.84

693.67

Freight and delivery

45.00

375.00

163.63

45.00

260.00

128.75

Indirect labor

Other indirect

31.50

125.00

86.03

50.00

100.00

77.86

Administration

Sales expense

Profit

250.00

807.00

526.41

370.00

613.00

491.00

Total, package and

erection

Direct material

2,017.07

3,696.96

2,970.43

2,017.07

3,448.00

2,709.01

Direct labor

947.44

1,647.29

1,253.48

947.44

1,647.29

1,262.28

Total direct costs

3,062.10

5,250.20

4,228.09

3,062.10

5,095.29

3,908.04

Freight and delivery

45.00

375.00

162.63

45.00

260.00

128.75

Indirect labor

65.36

604.39

183.82

111.63

185.38

143.27

Other indirect

50.00

659.18

327.78

50.00

659.18

361.50

Administration

29.94

197.00

97.30

29.94

52.28

39.88

Sales expense

37.57

259.98

151.38

152.40

259.98

200.41

Profit

450.00

1,192.46

865.55

450.00

1,192.46

842.13

Source: Office of the Housing Expediter figures.

146

Table 2 (Continued)

Summary of Unit Costs of Prefabrication Price Ranges

Wood

Metal

Element of Total Cost

From

Average

From

Average

Package, total

2,269.13

3,351.61

2,665.49

2,823.60

4,716.36

3,872.63

Erection, total

2,579.00

3,824.00

3,163.58

1,835.57

3,159.00

2,580.76

Total cost, less land

5,150.74

6,839.61

5,829.08

5,622.64

7,702.42

6,453.39

Package

Direct material

1,243.98

2,108.91

1,621.56

2,015.14

3,067.98

2,462.89

Direct labor

272.27

401.50

334.33

63.60

500.00

341.77

Indirect labor

65.36

65.36 *

72.00

604.39

304.41

Other indirect

116.30

1 16.30 l

126.82

534.70

270.34

Administration

82.00

142.56

112.28

117.45

197.00

147.23

Sales expense

90.77

123.50

107.14

37.57

172.96

124.47

Profit

226.91

335.16

266.55

306.40

471.63

407.19

Erection

Direct material

1,047.00

1,905.00

1,380.08

345.60

1,406.00

915.34

Direct labor

692.00

1,230.00

908.00

719.00

1,471.44

942.23

Freight and delivery

60.00

375.00

233.33

65.00

290.00

132.00

Indirect labor

Other indirect

75.00

125.00

108.33

31.50

75.00

60.50

Administration

Sales expense

Profit

465.00

646.00

533.83

250.00

807.00

560.00

Total, package and

erection

Direct material

2,666.47

3,507.91

3,001.64

2,415.60

3,696.96

3,249.65

Direct labor

964.27

1,613.25

1,242.33

1,058.60

1,535.04

1,259.20

Total direct costs

3,630.74

4,808.61

4,243.98

3,950.64

5,250.20

4,657.12

Freight and delivery

60.00

375.00

233.33

65.00

290.00

132.00

Indirect labor

65.36

65.36 '

72.00

604.39

304.41

Other indirect

191.30

191.301

126.82

609.70

305.84

Administration

82.00

142.56

112.28

117.45

197.00

147.23

Sales expense

90.77

123.50

107.14

37.57

172.96

124.47

Profit

722.17

981.16

800.38

721.63

1,189.00

967.19

1 Information on one firm only; therefore not an average figure.
Source: Office of the Housing Expediter figures.

347

Table 3

Breakdown of Cost to Consumer of Erected House without Lot

Total (%)

Plywood (%)

Element of Total Cost

From

Average

From

Average

Package, total

37.24

77.07

58.34

55.39

77.07

66.93

Erection, total

22.93

62.76

41.66

22.93

44.61

33.07

Total cost, less land

100.00

Package

Direct material

20.42

46.83

35.20

39.83

41.82

40.79

Direct labor

1.13

14.10

7.48

3.35

14.10

11.87

Indirect labor

1.07

7.84

2.89

2.05

3.65

2.55

Other indirect

1.89

7.68

5.20

6.36

7.68

7.14

Administration

0.64

2.56

1.53

0.64

0.69

0.67

Sales expense

0.56

3.44

2.46

3.20

3.44

3.37

Profit

3.72

7.71

5.92

5.39

7.71

6.93

Erection

Direct material

4.88

31.26

14.21

4.88

13.97

7.22

Direct labor

7.81

26.17

14.20

7.81

17.15

12.63

Freight and delivery

0.90

5.48

2.78

0.90

5.47

2.37

Indirect labor

Other indirect

0.48

2.43

1.49

1.09

1.82

1.38

Administration

Sales expense

Profit

3.82

14.36

8.84

8.15

9.23

8.53

Total, package and erection

Direct material

42.96

58.05

50.00

44.89

54.57

47.39

Direct labor

14.58

27.30

21.22

19.43

23.38

21.99

Total direct costs

61.19

78.21

71.72

67.77

74.00

70.20

Freight and delivery

0.90

5.48

2.78

0.90

5.47

2.37

Indirect labor

1.07

7.84

2.89

2.05

3.65

2.53

Other indirect

1.05

8.77

5.23

1.05

8.77

6.10

Administration

0.64

2.56

1.53

0.64

0.69

0.67

Sales expense

0.56

3.44

2.46

3.20

3.44

3.37

Profit

9.23

21.15

14.40

9.23

15.86

14.46

Source: Office of the Housing Expediter figures.

348

Table 3 (Continued)

Breakdown of Cost to Consumer of Erected House without Lot

tfood (%

Ketal (%'

I '

T?1 L f T* *. 1 /"* -

Element of Total Cost

From

Average

From

Average

Package, total

37.24

50.13

45.84

50.21

71.98

59.59

Erection, total

49.67

62.76

54.16

28.02

49.79

40.41

Total cost, less land

100.00

Package

Direct material

20.42

31.45

27.90

28.36

46.83

38.37

Direct labor

5.19

7.00

5.74

1.13

7.63

5.17

Indirect labor

1.07

1.071

1.07

7.84

4.17

Other indirect

1.91

1.911

1.89

6.94

3.70

Administration

1.43

2.34

1.89

1.79

2.56

2.20

Sales expense

1.49

2.15

1.92

0.56

2.64

1.88

Profit

3.72

5.02

4.58

5.40

7.20

6.31

Erection

Direct material

20.25

31.26

23.60

6.15

22.56

14.14

Direct labor

13.38

20.19

15.53

9.33

26.17

15.35

Freight and delivery

0.98

5.48

4.01

0.97

3.77

1.98

Indirect labor

Other indirect

1.23

2.43

1.89

0.48

1.33

0.93

Administration

Sales expense

Profit

9.02

9.45

9.15

3.82

14.36

8.83

Total, package and erection

Direct material

49.85

52.71

51.50

42.96

58.05

51.01

Direct labor

18.72

26.48

21.27

14.58

27.30

20.17

Total direct costs

70.30

78.16

72.77

61.19

78.21

72.59

Freight and delivery

0.98

5.48

4.01

0.97

3.77

1.98

Indirect labor

1.07

1.071

1.07

7.84

4.17

Other indirect

3.14

3.141

1.89

7.92

4.19

Administration

1.43

2.34

1.89

1.79

2.56

2.20

Sales expense

1.49

2.15

1.82

0.56

2.64

1.88

Profit

12.82

14.35

13.73

11.02

21.15

15.14

1 Information on one firm only; therefore not an average figure.
Source: Office of the Housing Expediter figures.

349

costs range from 35% to 45% of final price, less lot. This was a sub
stantially greater proportion than that shown by Table 3, in which
total direct labor costs accounted for 14-27% of the final price, less
lot, and total indirect labor costs accounted for 1-8%. The lower pro
portion of labor costs in prefabrication has been explained not only
by greater productivity, but also by lower hourly wage rates stem
ming from the lower skill requirements, better working conditions,
and steadier employment.

In considering the percentage allocated to sales expense, it should
be remembered that the data represented mostly new firms which
did not have established distribution systems. An unfortunately large
number of prefabricators during this period thought that, because of
the acute housing shortage, all that had to be done was to get the
production line moving— that somehow the process of getting the
houses from the end of the line to the customer's lot, financed and
ready for occupancy, was not a problem. Experience has proved
otherwise, and if similar estimates were to be submitted today, they
would probably include a much more substantial item to cover the
costs of establishing an organization able continuously to sell, finance,
erect, and service houses as they are produced. Once such an organi
zation was established and growing at a small but steady rate, how
ever, its percentage cost might well be reduced.

Another item that deserves attention is the sum of indirect and ad
ministrative costs. It may be true that these, as the above quotation
points out, "represent such small percentages as to offer little promise
of cost reduction." But to stop here would be to overlook at least two
important points. For one thing, while these costs may be a small
percentage of the total at high volumes, they may skyrocket as volumes
fall. During the past few years overhead costs have been the down
fall of more than a few newly established prefabricators who re
quired some time to smooth out their operations and who, by the time
they had overcome the problems of marketing, found that their
working capital had been consumed in such expenses.

More important is the relationship between the overhead encount
ered when the building process is moved into a factory and the sav
ings in labor cost thus achieved. Clearly, from the production stand
point, if the additional indirect expenses outweigh the savings in di
rect costs, it is uneconomical to shift an operation from the field to a
plant removed from the site. This point has been very well sum
marized by Robert W. McLaughlin, a veteran prefabricator:

Criticize the so-called construction industry as you will, it has demon
strated its ability to operate in the field at an extremely low overhead.

350

Exclusive of insurance and social security charges, overhead on construc
tion labor is of the nature of 5-10%. In any factory, on the other hand,
overhead on direct labor will vary from 100% to 300% or even more.
My own experience with wood fabrication was that factory overhead ran
something over 100%. That is factory overhead only, without adminis
tration or sales expense. I am told that a plant of average mechanization,
such as a vacuum cleaner plant, will have an overhead ratio to direct labor
of about 150%, and that in more highly mechanized straight line produc
tion the rate will be of the nature of 200% or even higher. What does
this mean with respect to the factory processing of wood? Assume a field
labor operation costing $100. With 10% overhead the operation per
formed in the field will appear on the cost sheet at $110. Along comes
the prefabrication enthusiast who assures you that he can save 40% of the
direct labor cost by doing it in the factory— that he can do the $100 opera
tion in the factory for $60 worth of labor. 40% is quite a saving. But
immediately he has to add at least 100% factory overhead, and his true
cost becomes at least $120 as against a field cost of $110. Also we have
to think about additional transportation and handling. It is apparent
that removal of a labor operation from the field can be justified only if the
direct labor saving is really great— of the nature of 75% or 80%. This
substantiates our earlier statement that if we are to change the locale of the
process at all we have to change the overall process radically. We also
categorically state that the nature of wood does not present enough oppor
tunity for mechanization to warrant a shift in the process from field to
large, central factories.43

This telling comment by one who has spent more than 15 years
prefabricating in wood, metal, and other materials is not to be
brushed lightly aside. McLaughlin's estimate that factory overhead
costs amount to about 100% of direct labor costs is substantiated by
the figures in Table 4. In this breakdown it can be seen that direct
factory labor and factory overhead are roughly equal. In very few
circumstances have prefabricators yet achieved savings in direct labor
of 75-80%, and, in the light of the above reasoning, this may offer at
least a partial explanation for the somewhat disappointing results of
prefabrication in cutting the cost of building to date.

2. Budget Cost Figures of a Large Producer of Stressed Skin Plywood
Houses

Table 4 presents the percentage breakdown of unit costs for the
package only. The figures indicate allocations of cost expected in
order to break even on an annual production of 1,500 units, with the

43 Talk delivered at Massachusetts Institute of Technology, February 26, 1948.

351

indicated net income serving merely as a safety margin. At least 10%
profit would be required for a continuing operation. Production of
more than 1,500 units would lower percentages for plant expenses
and for sales, general, and administrative expenses. As these were
lowered through increased volume, the gross profit and net income
would increase accordingly. The house in question was of stressed
skin plywood construction and was being produced in one of the best-
equipped plants in the industry.

Table 4

Budget Cost Figures Based on 1,500 Houses per Year

A Large Manufacturer of Stressed Skin Plywood Houses l (January 1, 1948)

Item Per Cent

Total house package (f.o.b.) 100 . 00

Direct materials

Processed materials 52.91

Jobbed materials 24.75

Total 77.66

Total labor

Direct manufacturing 3 . 82

Rework and repair 0.36

Materials handling and shipping 1 . 96

Service and maintenance 0.72

Wage premiums 0.06

Total 6.92

Margin above materials and labor 14.43

Indirect plant expense (materials and service) 2.24

Plant overhead and administrative expense 3 . 89

Total plant cost 91.70

Gross profit 8.30

Sales, general and administrative expense

Selling expense 2.02

General administrative expense 2.63

Total 4.65

Operating profit 3.65

Other income and deductions (net) 1 . 48

Grand total all costs 94.87

Net income before taxes 5. 13

1 These figures are for a 24' X 28' house. Package price, f.o.b., $4,100. The average
price of this house, erected but less lot, would be about $7,000.

352

Table 5

Cost Breakdown for TVA Sectional House (1943) x

Dollars

Item Per Cent per House

Gross sales 100.00 2,673

Materials 43.96 1,175

Labor 18.71 500
Plant burden

Rent 0.94 25

Heat 0.56 15

Light 0.56 15

Power 0.56 15

Insurance 0.56 15

Maintenance 0.56 15

Supervision 2.24 60

Cost of manufactured goods 68 . 65 1 , 835

Manufacturing profit 31.35 838

Expenses

Selling 2.43 65
Shipping

Loading 0.56 15

Weather protection 0.56 15

Trucking 5.16 138

Permits 0.37 10

Unloading 1.31 35
Field assembly

Labor and materials 2.24 60

Supervision and overhead 1.88 50

Advertising 0.37 10

Administration 2.62 70

Social security and taxes (except income) 1.68 45

Total expenses 19.18 513

Operating profit or profit before depreciation 12. 17 325

Depreciation 1.88 50

Net profit before federal taxes 10.29 275

Federal taxes 5.61 150

Balance 4.68 125

Interest on invested capital 0.94 25

Net profit on sales 3.74 100

1 Erected house less furniture and equipment (range, refrigerator, water heater, and

space heater), and excluding land, foundation, and site utilities. Two-bedroom house,
24' X 24', three sections.

Source: Estimates by TVA which were reconciled with the experience of several firms
having contracts for production of these houses.

353

On Table 4 it will be noted that factory labor costs were a very
small part of the total package cost and that materials represented by
far the biggest item. This is partly due to the fact that the materials
as they were received had been largely cut and milled to size, and
the factory operations were chiefly assembly and finishing. It can also
be seen that factory overhead was somewhat greater than direct labor
costs, but it is necessary to consider that in this case some of the fac
tory overhead was expended on the storage and handling of finished
materials and equipments which were included in the package sent
to the dealer and should, for this reason, have been allocated to the
dealer's cost sheets rather than to those for the manufacturing opera
tion. The sales expense represented only the prefabricator's costs in
this breakdown and did not include expenditures by dealers.

Table 6

Cost Breakdown for AIROH House

British Temporary Housing Program (1947 Estimates)

Item Per Cent Pounds (£)

Production

Materials, fixtures, and fittings 51.8 847

Factory fabrication and assembly 17.0 278

Other production costs 2.7 44

Factory plant and equipment 2.6 43
Transport

Expenditure on vehicles, spares, and repairs 1.5 25

Haulage 2.6 43

Grading, utilities, and foundation 14.6 238

Erection 3.3 53

Contingencies 1.7 28

Overhead costs 2.2 36

Total 100.0

Less net residual value of productive assets

£1,610

Table 5 is a cost breakdown for a TVA sectional house. The house
measured 24' X 24', had two bedrooms, and arrived at the site in
three sections. The figures are for 1943 and are based on estimates
by TVA which were reconciled with the experience of several firms
having contracts for the production of these houses. The principal
point of interest here is the extent to which the manufacture of the
house had been transferred to the factory. The motive behind this
was probably more the desire to reduce site labor requirements than
it was economy. Site labor had to be kept to a minimum because of

354

the wartime shortage of construction labor, because of the desire to
reduce the number of people and the confusion at the site, and be
cause of security reasons, since about 5,000 of these sectional houses
were built at Oak Ridge, Tenn., and Hanford, Wash., two of the
atomic energy production centers. The relatively large shipping costs,
about 8%, were due to the fact that transporting the sectional house
involved careful handling of a finished product, which included much
empty space, over relatively large distances in some cases.

Table 6 is a breakdown for the British AIROH house, a sectional
aluminum structure which was produced in large quantities in aircraft
plants after the war. The figures are estimated rather than official, but
they serve as an informative basis of comparison with the figures for
the sectional wood TVA house. It will be noted at once that, because
this house is sponsored by the government, advertising and selling are
not items of cost.

355

II.

Part

Chapter

MARKETING

I. Introduction

The marketing aspects of any industry are properly defined as in
cluding "all business activities involved in the flow of goods and serv
ices from physical production to consumption/' * For the pref abrica-
tion industry, this includes the determination of markets, prices, chan
nels of distribution, and methods of sale; and the procedures used in
financing, site selection, transportation, erection, and servicing. Many
of these subjects have been discussed in earlier chapters, for market
ing considerations have an obvious influence on decisions regarding
production, procurement, design, and management, although the ex
tent of this influence has not always been recognized in the industry.

In the period immediately following the war, the breadth and im
portance of marketing problems were not generally appreciated. Pro
curement and production problems were far more pressing, and, with
the demand for housing running at the highest level in recent history,
it was easy to visualize an eager line of customers, checkbooks in hand,
waiting to claim the houses as they came from the plant. Few of the
companies in the field had had any experience selling prefabricated
houses; many had never sold houses of any sort. Furthermore the
industry was young, the war had been won, and it was not hard to
dismiss as gloomy conservatism the warnings of those who had learned
about marketing the hard way during the depression.

During the period of the survey, the marketing lessons were gradu
ally being learned. Government contracts terminated, and bidding
for large projects began to mean cutting costs and profits to the bone.
High hopes engendered in the days of the Veterans' Emergency Hous
ing Program began to dissolve, and slowly the real bottleneck was lo
cated—at the end of the assembly line. Foster Gunnison, who had
always placed marketing first in order of importance, had warned
the industry in 1944:

It is obvious that orders must flow into the plant, each day, at the same
continuous rate the houses flow off the conveyors. . . . The investment in
a mass-production plant is so great that it will only pay-out by keeping the
plant going to capacity every day. To provide a continuous flow of orders,
therefore, becomes the most important problem of all. Thus, upon the

1 Harold H. Maynard and Theodore N. Beckman, Principles of Marketing ( 4th
ed., New York: Ronald Press, 1946), p. 3.

359

method of distribution and sales used, depends the ultimate success or
failure of the industry as a whole and each company within it.2

Nearly ten years earlier, John Burchard had been even more precise:

The focus of efforts so far has been on the redesign of the structure of
parts of the house, often very ingeniously. But the trouble with these ef
forts has been that they run squarely against the stone wall of the amount
of capital required to bring an old un-mass-produced product into mass
production almost over night, and the economies proposed are available
only if the mass production is achieved. A sounder approach, it would
seem to me, might be made by regarding the problem at the outset as one
of marketing. After marketing success with a semi-orthodox product, the
economies and advantages of new structures might be incorporated.3

Marketing patterns were being formed at the time of the survey,
in many cases very elementary, in a few cases more advanced, and
the rest of this chapter is devoted to describing these patterns.

II. Markets

A. Market Areas

The prefabricated choice of market areas was greatly influenced
by the type of product he wished to offer and by the manner in which
he wished to offer it. If he decided to make a complete and distinc
tive house, bearing his trade name, he would usually plan to sell it
either in large urban centers in direct competition with the operative
builder, or in rural areas where there were fewer problems with codes,
labor unions, and competitors. If he preferred to make a factory
package, to be put into the final house without identification of the
maker, he would usually plan to sell it either to large speculative
builders in the cities or to small contractors and individuals spread
over a wide area. The preference of the prefabricators with refer
ence to a few simple classifications of market areas, and the reasons
which they gave for their choice, are summarized below.

2 Foster Gunnison, "The Economics of Mass-Distribution and Mass-Sales of
Prefabricated Homes," Prefabricated Homes, 2 (February 1944), 23.

3 Burchard, "Prefabricated Housing and Its Marketing Problems," p. 154.

360

T. Metropolitan Areas

An almost exclusive interest in the metropolitan areas, roughly de
fined as those having populations of 100,000 and upwards, was ex
pressed by 25 companies. On statistical grounds alone this would
have been a good choice, since census figures indicate a continuing
trend in the United States for the population to move into such areas
(and, within them, to move outwards from the built-up centers of
cities ) . Despite the fact that the metropolitan areas had a somewhat
smaller proportion of single-family houses than the rest of the country,
they probably contained almost as large a total number of such
houses. While the built-up centers were characterized by high land
costs and stringent building restrictions, even there certain prefabri-
cators felt they might have advantages to offer. For example, the
fireproof house built by Fabcrete of America, Inc., could be erected
in districts from which wood frame houses were excluded in the
interests of fire prevention.

Most of these 25 companies, however, were interested in the sub
urban fringe, which offered such attractive features as wide selection
of building land at suitable price relatively close to a concentrated
demand, relatively broad range of demand, convenience of transporta
tion, likelihood of many vacant lots already provided with streets and
utilities, and the best general prospects for large projects, whether to
be built for sale or for rental investment. Particularly for those who
produced unconventional houses, the concentration within metropoli
tan areas of young business and professional families and of families
of relatively high incomes was a decided advantage.

2. Smaller Urban Areas

More desirable to the average prefabricator, despite the advantages
of metropolitan areas, were the smaller urban areas, where the popu
lations ranged from 2,500 (the smallest urban area in census compu
tations) to 100,000. In all, 52 companies expressed a preference for
such market areas, with the major interest in the more populous areas
within this range. The prefabricators mentioned several special ad
vantages in such areas. They were generally considered to have
lower wage scales and other operating costs; this made them low-cost
plant locations, and low cost meant broad marketing advantages.
While the overall demand was not so large as in a metropolitan area,

361

it was nevertheless adequate in view of the scale of operations of
the average prefabricator, as was the available supply of building
sites. The costs of improving the land were not so great in smaller
cities, where standards were usually lower, development less inten
sive, and wages and costs lower. Taxes almost always were lower in
smaller cities than in metropolitan centers, although metropolitan
suburbs might compete on this score. It was usually considered
easier to establish friendly relations with trade unions, with the vari
ous municipal departments, with bankers, and with potential cus
tomers in the smaller cities. The advantages of speed and efficiency
offered by a dealer in prefabricated houses were found to be rela
tively more apparent in the smaller cities where large-scale builders
were rare and therefore the dealer had a relatively better risk in the
use of his capital. And, finally, the aggregate of orders flowing in
from a diversified selection of smaller cities where these favorable
conditions might be found was considered to yield a steadier rate of
production than would be the case with orders flowing from any one
metropolitan area.

In the very small urban areas these arguments lost some of their
force. The tendency of the population to move towards the cities
meant that demand for houses was often less in the smaller towns; the
inhabitants were noticeably more conservative in their tastes and in
their manner of doing business; and because the prefabrication plant
was itself likely to be located near a somewhat larger city, transporta
tion costs were often higher.

3. Rural Non-Farm Areas

A preference for the rural non-farm area, defined as including com
munities of less than 2,500 population which contain little land in
farm uses, was expressed by 22 companies. From census figures this
would seem to be by far the best market for prefabricators, since
almost as many total single-family dwelling units are being erected in
rural non-farm areas as in urban areas.4 There was a very real feel
ing on the part of many of the prefabricators that this constituted
their best market. John Richardson, whose experience lies in financ
ing and sales, told those attending the December 1947 PHMI Winter
Meeting that in his opinion the "market is 75% in rural areas and small

* Construction, U. S. Bureau of Labor Statistics (January 1948), p. 4.

362

towns," and C. W. Farrier, formerly head of the Technical Division
of NHA and more recently research director for Gunnison Homes,
earlier had said almost precisely the same thing.5 The advantages
of such areas lay in the possibility of erecting a good-quality house
with very reduced site labor requirements on a site far removed from
the nearest skilled conventional builder. Frequently houses designed
for these areas would have less costly finish and equipment, although
the prefabricator might well attempt to include within his package
as much as possible of the necessary materials and equipment. Fre
quently, also, the houses were to be designed for minimum site im
provement ( probably without a basement ) , and for erection processes
involving as little special equipment and skilled labor as possible.
The design of the house itself could be highly standardized since it
would not be frequently reproduced within the area. And perhaps
the largest single factor favoring the rural non-farm areas was the
fact that conventional builders in such areas were at the very end
of the normal materials distribution channels. This gave the prefab
ricator, with his greater buying power and speed, a very decided ad
vantage. Admittedly, the establishment of suitable sales methods and
distribution forces to reach so scattered a market offered a difficult
problem, and it was one which no prefabricator had fully solved at
the time of the survey, although many were keenly interested in the
possibilities offered by so broad and stable a potential market.

4. Rural Areas

Eleven companies indicated a preference for rural areas as a
market for their houses, such areas being defined as those devoted
primarily to farming. Most of the companies featured houses which
could be erected by the farmer himself, who represented one of the
few groups in the consumer population generally capable of doing an
efficient job of erection. Many of these farm cottages are highly
standardized in design, the usual theories about the need of apparent
variation being dispensed with in view of the wide scattering of pur
chasers. Frequently companies operating in rural areas also prefab
ricated farm utility buildings; indeed many entered upon the prefab-
rication of houses from that field, for example, Pre-Fab Industries
Corporation and Economy Portable Housing Company.

5C. W. Farrier, "Prefabrication in Post-War Housebuilding," Prefabricated
Homes, 2 (February 1944), 11.

363

A specialized form of the rural market was the market for recrea
tional cottages, sought after by an increasing number of prefabricators
after the lifting of the restrictions of the Veterans' Emergency Hous
ing Program. Hodgson, probably the oldest continuing prefabricator
in the business, had been making a large share of its sales in this
market since 1892. Here designs commonly varied widely with de
mand, and houses could be greatly simplified by the temporary and
usually warm-weather nature of their intended use. Structurally, the
houses were generally panelized into sections capable of being easily
manhandled, and the erection system was usually simple enough to
permit the use of unskilled labor on rather rough and isolated sites.

B. Special Market Types

Prefabricators had varying preferences with regard to channels of
distribution; in the selection of these channels, they were often also
making a choice between two broad types of market: that in which
distinctive houses, given a sort of "brand name" by advertising and
promotional efforts, were sold to the public; and that in which special
ized house packages, varying according to the circumstances involved
in the order, were sold to the dealer or builder who offered them to
the public without announcement of the identity of the fabricator
of the basic package. Of the former, Lustron was a good example,
and of the latter, American Houses. In addition to this basic distinc
tion in market approach, several special types of market deserve
further discussion.

1. Industrial Markets

At least 15 companies concentrated a major part of their efforts
on selling large groups of houses to industries building for their em
ployees. This was a natural outgrowth of the war period, during
which sales had been made to government agencies in large quantities,
and of the period of boom construction immediately following the
war, when new housing was needed near new plant facilities. It
was easy for the prefabricators to shift over from large government
orders to large industrial orders. American Houses sold units to the

364

builders of several such projects, among them one for 250 families
in Manville, N. J., to house employees of a Johns-Manville plant.
U. S. Homes developed special low-cost designs adapted to the needs
of southward-migrating textile companies. Nygaard Builders, Inc.,
developed for a Pittsburgh contractor a unit designed for housing in
coal-mining communities. With the decline in postwar industrial ex
pansion and the general leveling-off of business activity, this market
was showing signs of shrinking, but during the period of the survey
it still was a significant factor in the plans of these prefabricators.

2. Export Markets

Among the companies interviewed, six indicated that they had
shipped houses outside the continental limits of the United States,
and seven more said that they were making definite plans in that
direction. Other companies expressed interest, but had no plans at
that time. Aside from the lend-lease program,6 however, actual sales
in foreign markets had been small, and such sales as there were came
about as the result of special circumstances rather than any serious
demand on the part of foreign consumers.7 This was, of course,
partly the result of the dollar shortages in most of the potential con
sumer countries, but partly also it reflected the difficulties and costs
inherent in purchasing houses in the United States and shipping them
abroad for erection and use under what often were very unfamiliar
conditions. Companies seriously interested in the export market soon
realized that special models, involving a considerable degree of re
design and the changing of dies and jigs, would be required, and that
in most cases the redesign would have to be in the direction of simpli
fication.

Transportation costs, when added to the high costs of production
in the United States, constituted a serious difficulty. Unless extra
handling and shipping costs were to be incurred, furthermore, units
would have to be designed so as to permit their being broken down
for shipment into relatively light and small packages.

For a time, during the worst of the materials shortages, government
quotas were a further limitation on the export business. Quotas under
the Second Decontrol Act, for instance, lumped prefabricated wooden

6 Chapter 2, p. 60.

7 Office of International Inquiries, HHFA, in an interview June 4, 1948.

365

houses with other wood mill products, and the unit limits were set as
follows:

Year Quota

1946 Closed. Each applicant examined individually.

1947 1,150

1948 3,440

However, such were the general difficulties that even these limited
quotas were never filled. For example, during 1948 the quota was
3,440 houses, export licenses were issued for only 1,697, and only 330
were actually shipped abroad. These quotas were more recently
entirely lifted, and it would have been possible to develop a good
export business under certain conditions if more foreign countries
had favorable dollar balances. As it was, the new country of Israel
was nearly the only one able to devote dollars to housing, and Israel
could not afford to spend its dollars on prefabricated houses designed
and equipped for living patterns in the United States.

More likely to be shipped abroad have been machinery, materials,
techniques, and skilled technicians. Six of the companies interviewed
had exported their "pattern of operations" in whole or in part. This
was particularly true of the sponsors of systems for the production
of concrete houses. Wallace Neff, for example, reported for Airform
Construction a Mexican licensee building schools in Mexico City and
houses in Acapulco, a Brazilian licensee with houses under construc
tion, and contracts or negotiations for contracts under way in Vene
zuela, South Africa, India, Egypt, Morocco, Spain, and Portugal.8
Others also were involved in this way: Precision-Built Homes had a
licensee in Canada and was considering arrangements for others in
South America; Soule Steel had developed a special house for the
Hawaiian market, only the steel parts of which it planned to export.

It seemed likely at the time of the survey that the major purchasers
of actual houses exported from the country were likely to continue
to be United States companies operating abroad. For example, in
1947, all the 275 wood prefabricated houses exported had such desti
nations: 180 went to a United States business firm in the Dominican
Republic, 40 to the Saudi Arabia Oil Company, and 55 to other identi
fiable commercial customers.

8 Interview in Los Angeles, Calif., April 16, 1948.

366

III. Pricing Policies

* Some honest confusion has usually attended any discussion of the
selling price of prefabricated homes, for prefabricators offered many
different kinds of prices. The lowest possible quotation was for a
house package f.o.b. factory, but some quoted the cost of the house
package plus transportation to the site. More commonly it was the
erected price, less the cost of the land, although in a few cases the
price included the land upon which the house was erected. In nearly
all cases, some extra features were included in the price, such as
built-in furniture, completely installed bathrooms, or kitchen ap
pliances.

The pricing structure can conveniently be examined in terms of the
experiences of 12 companies that were studied in 1947 by the
Flanders Committee.9 Cost data from the report of that Committee
were given in detail in the chapter on production; selected data are
reproduced here, with the prices of all the companies averaged to
gether to give a representative picture. Using the total cost of the
erected house, less cost of land, as 100%, the following relationships
were significant:

Average house package $3 , 460 .67 58 . 34%

Average cost of erecting the house 2 , 448 .27 41 . 66%

Total cost, less land $5,908.94 100.00%

Looking further into the erection costs, it is found that $162.63, or
2.78%, was made up of freight and delivery costs. A combined profit
was taken on the package and erection of $865.55, or 14.40% of the
selling price. However, $350.70, or 5.92% of the selling price, repre
sented profit attached to the house package, and thus went to the
manufacturer.

Obviously, the price to the ultimate consumer was far more than
just the house package cost. That, nevertheless, represented a feas
ible starting point for an examination of prices during the survey
conducted by the Bemis Foundation. In the winter and spring of
1946-1947, 54 companies offered house packages at an average price
of $4.02 per square foot. The average erected price of these houses,
usually as quoted by them, was $8.45 per square foot, exclusive of

9 High Cost of Housing, p. 150.

367

the cost of land, and the average size was 762 sq. ft. The average
selling price which resulted, $6,439, was probably lower than the
typical selling price for the industry, because many of the 54 com
panies surveyed tended to have lower than average prices, and geo
graphic variations in costs and in quality standards made a difference.
The period studied was one of advancing prices, so that figures more
recent than these would be considerably higher for comparable
houses.

Further information on selling prices is furnished by a PHMI sur
vey of its membership made in 1947. The median price for prefabri
cated houses was then found to be $7,000, exclusive of land, with
prices ranging from $5,100 to $8,000. The most common size of
house was 24' X 32', or 768 sq. ft.

By way of comparison, the average construction cost per unit of
all single-family dwelling units started in the country was $5,525 in
1946, $6,750 in 1947, and $7,850 in 1948.10 It should be emphasized
that these figures, while including an allowance for builder's profit,
do not include the cost of land; they represent only construction costs
and not selling prices.

In general the price policy of a house manufacturer seems to have
been determined by applying to his production costs an average
markup selected to yield a reasonable profit for him. There was very
little tendency on the part of the manufacturer to charge a price out
of line with a fair return; the realization of the need for mass sales
in order to maintain steady production seems to have served as a
curb on his desire for immediate profits.

The price policy of dealers generally was a somewhat different
matter. Dealers followed no single pattern, but a large number were
inclined to take full advantage of the seller's market then prevailing,
with little thought to future sales volume. During the period when
price control was in effect, the OPA allowed a 10% dealer's markup
on the cost of the house package. Prefabricated Homes magazine,
stating the case for a higher markup, estimated that overhead ex
penses would amount to 7% of the selling price.11 If a 7-10% net profit
to the dealer were added to this, the resulting markup would be
around 15%. Those companies which attempted to limit the dealer's
profit usually allowed a markup of 15-20% of the sales price. Most
franchises, however, gave the dealer the authority to set his own sell
ing price; in most cases they had to, if for no other reason than varia-

10 Housing Statistics, Housing and Home Finance Agency (May 1949), p. 4.
Bureau of Labor Statistics figures.

" Prefabricated Homes, 6 (May 1946), 9.

368

tion in local land and improvement costs and wage rates. Many
companies felt that better control of dealer prices in the future was
essential to the industry, since one of the purposes of prefabrication
was to provide the ultimate consumer with housing as good as or
better than conventionally built structures, at a lower price.

The nature of the competition offered by conventional builders and
other prefabricators determined to some degree the dealer's price
policy, especially when dealers were located in areas where competi
tion for the housing customer was becoming more severe. In par
ticular, those dealers handling a fairly conventional house seemed to
follow the price leadership of the operative builders in the area.

An exception to the general price averages was found in the case
of 14 companies which concentrated on the higher price and more or
less custom-design market, and commonly made use of modular panels
or job-lot order modular component production systems. Sometimes
the theory of such prefabricators, as presented by George Fred Keck,
designer of Green's Ready-Built solar house, was to build for a quality
market, relying on the fact that the factory can put on better finish
and detailing than can be reproduced by a local building contractor
at anything like comparable cost. A greater value, rather than a lower
delivered cost, was the object, and it was hoped that, when this
market was established and production costs cut, still more equipment
and better value would be added instead of reducing prices for the
consumer. Presumably, such houses would be particularly well de
signed to appeal to the income group which might ordinarily hope to
have a very small house hand-tailored by an architect. With this
quality market established, such a company might then consider
bringing out a lower-priced model for a broader market, relying on
the advertising appeal gained by its more expensive models.

On the other hand, a few companies were interested in producing
austere shelter in the very low price ranges. For example, the Texas
Housing Co. was selling its "Homette," a 16' X 16' plywood cabin,
for $463.24, knocked down, early in 1947 (see Figure 44). The
Wingfoot house, an expandable trailer with an area of 256 sq. ft.,
was being sold for as little as $3,000, ready for occupancy. Somewhat
above this level, many prefabricators were starting to manufacture
houses of standard size and equipment priced at $6,000 or less, erected,
but excluding the price of the land. These structures had no base
ments, but standards of design, material, and construction were at
least as good as those of similarly priced conventional houses. In
1948 National Homes Corporation brought out a two-bedroom house
to sell for $2,089 f.o.b. the plant. With added costs of erection, wir-

369

ing, plumbing, etc., the two-bedroom model sold complete with lot
for only $5,750, and could be purchased with a $300 down payment
and monthly payments of $34.87. There was also a three-bedroom
model priced with lot at $6,150 (see Figure 46).

Each model had a living room, a bath, a utility room, and a kitchen
equipped with built-in cabinets, counter sink, and a laundry tray.
Included with the house were an oil heater in the living room and an
automatic water heater in the utility room. Plywood was used for
the exterior finish, and inside walls were of waterproof, crackproof,
room-size Upson board, used in natural finish, painted, or papered.
The houses were erected on an insulated concrete floor with no base
ment and with no doors on the bedroom closets.12

After the initiation of the Economy House Program by the Hous
ing and Home Finance Agency in cooperation with the building in
dustry, most of the other companies announced special low-price
models (see Figure 47). One, for example, designed by General
Industries, Inc., to sell in the $6,000 bracket, was described by PHMI
as follows: ,

This economy home is a ... one story model with two bedrooms, liv
ing room with dining area, kitchen, bath and a utility room, with additional
storage room in the attic space. It is 24' 3" square, of stressed-skin plywood
construction and is erected on a concrete slab. The inside walls are finished
with wall paper or may be painted. . . . The buyer is offered several
choices of exterior finish.

Approved by the FHA for mortgage insurance, the houses are being
financed under the new provision of the Housing Act of 1948 authorizing
government-insured 95% mortgages on owner-occupant homes where such
loans do not exceed $6,000.18

On the whole, however, it is fair to say that the prefabricated hous
ing industry was only beginning to produce houses at a low cost for
the mass market. Indeed, many prefabricators did not feel it should
try to do so. One writer stated even before the postwar rise in prices
that new houses should be priced from $6,000 to $8,000 rather than
from $2,000 to $4,000, because the effective buying power resided
in the seven or eight million families who represented the upper 20-

12 National Homes estimated that 90% of the 25 houses per day produced
throughout most of 1949 were these "thrift homes." Many features of equipment
and finish have been added to recent models without increase in price.

13 PHMI Washington News Letter, September 24, 1948, p. 3. By the time of
the PHMI Fall Meeting at Winnipeg, October 1949, practically all member com
panies had come out with "economy" or "thrift" homes, and such homes repre
sented 75-80% of total production in several cases.

370

25% of the income group.14 And the $6,000 house of 1946 would cost
close to $8,000 in 1948. Many outside the field felt that undue concen
tration on cost was producing houses of dangerously low space stand
ards.

A few shared the feeling, best expressed by Carl Strandlund of
Lustron, that it could not be expected that a family in the low-income
group would be happy to invest all its resources in what was loudly
proclaimed as a bare minimum house; it would prefer to pay a little
more and get some extras— some genuine "quality" features— that
would give a real pride of ownership. This feeling led Strandlund
to invest money in top-notch architectural services for the overall
improvement of the Lustron house in future models, and it also was
behind the production, in 1950, of a three-bedroom model containing
1,209 sq. ft., and of garages, for one or two cars, which could be con
nected to the houses by breeze ways.

On the other hand, the market pressure was such that Lustron, too,
was prepared just before its failure to bring out an economy line.
The Lustron Newport homes took full advantage of standard parts,
running the regular roof trusses across the long dimension of the
house to avoid the production of new structural members. There was
to be a two-bedroom model containing 713 sq. ft. and a three-bed
room model containing 961 sq. ft. at prices competitive with the
economy lines which made up almost the entire output of Lustron's
competitors.

Another point of view was expressed by William K. Wittausch:

Even though families move in order to improve their housing condi
tions, they need by no means move into new houses as evidenced by many
millions of families who today live in houses which were not newly built
for them but which are better than the houses they left. That is why the
housing needs of millions of families, especially in the low-income group,
do not necessarily represent a vast potential market for new, low-price,
mass-produced houses.

. . . Whether new or old, the quality of housing a family is able to
occupy depends almost exclusively on its income. ... it is only natural
that the higher income families move into newer and more desirable houses
first, with the families that cannot afford the pleasure of moving into a
fresh, new house moving into the older and less attractive existing dwell
ings left by those who move out. New prefabricated houses like other
new houses automatically command the same premium for freshness re
gardless of the price group in which they are offered. It would appear,
therefore, to be more advantageous to prefabrication if the current em-

14 Neal MacGiehan, "The House for the Mass Market," Prefabricated Homes,
5 (February 1945), 16-7.

371

phasis on producing low-cost houses rather than on putting higher value
into houses relative to other new houses were to be reversed.15

Other prefabricators have been frank to state that the industry
cannot produce new homes for the lowest-income group,16 and that it
should recognize the need of public housing for that group and the
related possibility of a firm government housing policy to which pre
fabricators might adjust intelligent plans for operation under settled
conditions over a long period.17

It is not the province of this discussion to attempt to find a method
of meeting the need for really low-cost housing for the low-income
groups. The survey indicated that as yet the prefabricated housing
industry had not come up with the solution to this problem. But the
problem was recognized, and efforts were being made by most pre
fabricators to lower costs. In the back of their minds seemed to be
the hope of capturing the mass market with a house that cost no more
than present secondhand houses, and yet was superior to them in most
respects.

IV. Channels of Distribution

The pattern of handling goods between production and consump
tion, the channel of distribution, is determined by the system of hand
ling and storing the components, the method chosen for moving the

15 William K. Wittausch, "Marketing Prefabricated Houses," Harvard Business
Review, XXVI (November 1948), 696-7.

16 It should be added that several prefabricators believed, with Fred Gentieu
of Plainfield Lumber & Supply Co., that they could reach the lower price ranges
only in units other than single-family detached houses, that is, in row house or
apartment units.

17 While the industry has been officially opposed to individual public housing
bills in the past, such views as this have been expressed by some of its most
thoughtful members. In his address to the PHMI membership in December
1947, John C. Taylor, Jr., President, American Houses, Inc., said: "The people
in this country are going to be adequately housed, and if private industry does
not supply this housing, it is going to be supplied through Government subsidy.
. . . The majority of you do not like subsidized housing any more than I do,
but yet, if we are really true to ourselves and will bring our innermost convic
tions to the surface, we know that that statement is true/'

372

goods at low cost, and the middlemen selected. No single system of
distributing houses was common to all prefabricators, and several
companies employed more than one channel.

A. Factory Direct to Consumer

From factory to consumer is the most direct method of distribution.
In some cases manufacturers employing this channel made the erec
tion of the house the consumer's responsibility, while in others the
manufacturer himself took care of the erection.

1. Erection by Purchaser

Only one company sold all its houses for erection by the purchaser
himself, but 19 companies sold part of their output for such erection.
The simplicity of this distribution scheme appealed most often to the
newer companies, especially those on the West Coast. However,
while it is true that such a scheme was simple, it often involved the
drawback of a specialized or limited market. Financing requirements,
and the small number of customers willing to be responsible for erect
ing a full-size house, were the chief limiting factors. The FHA was
reluctant to approve loans based on purchaser erection, and so this
scheme usually was limited to companies offering non-FHA minimal
units. Allied Building Credits was willing for a while to grant loans at
high interest rates on such unpredictable risks, but this specialized
financing firm soon became inactive in this field.

Immediately after World War II a large number of such units
were produced as prefabricated garages. These garages, usually
two-car size, were purchased by veterans in desperate need of hous
ing. Nicoll and Co. sold 20' X 24' panelized cottage shells for $792.
The John L. Hudson Co. produced as many as 80 garages a day,
probably 50% of them used as dwellings.

While only a small percentage of the total housing market was will
ing to take the responsibility for erection in return for potential sav
ings in cost, it seemed likely that there would always be some who
would prefer this method. These purchasers liked the convenience
of buying most of the materials for a house in one package, and were
glad of the chance to jeduce cash outlay by contributing their own

373

labor. Most of these purchasers were farmers, veterans, and build
ing tradesmen, often operating within the framework of a coopera
tive. Some, however, like the purchasers of vacation cottages, were
interested more in the convenience of getting delivery of a unit of
known quality at a remote site, and in shortening construction time,
and they were not likely to realize substantial cost savings.

Within the industry, it was generally considered risky to sell units
direct to private owners for erection, and there was a growing senti
ment that the prefabricator should assume the responsibility for see
ing that the agent of erection performed the building operation in a
satisfactory manner. Unless the house were so designed that erection
became nearly as simple as connecting up a trailer, many prefabri-
cators felt that savings inherent in good organization of site work
might well be lost by purchaser erection. One prefabricator stated:

The prefabricator who will stay in business will furnish a complete
house, key in door, at a fixed price, and will be responsible for erection
and finishing. The days of shanty jobs are over; the days of shell building
are drawing to a close.18

1 In two different patterns, however, this channel of distribution was
well established. The precut house, as produced by Aladdin and
many others,19 typically was distributed in this manner, and had
been for 40 years. Indeed, Aladdin had tried a system of dealer-
erectors some 12 years before the survey and had decided that the
direct mail-order business, with individuals acting as their own build
ing contractors, mostly in rural non-farm areas, was better suited to its
purposes.

On the other hand, factory sales organizations frequently sold large
groups of houses to a contractor, a municipality, or an equity in
vestor. This middleman then went on to erect or to make the arrange
ments for erection by a contractor. Dealers were ordinarily not so
well able to handle such sales, and some companies reserved the right
even in exclusive dealer franchises to make sales of this sort them-

18 C. F. Dally, President, Prefabricated Products Co., Inc., interviewed January
21, 1947.

19 Sears, Roebuck and Co., which had sold precut houses from 1911 through
1942, brought out its Homart house in 1947. This was designed as a ready-to-
erect house, partially precut, partially prefabricated, and partially of random-
length materials to be cut to fit in the field. Sales were made through mail
order catalogues in Philadelphia, Boston, Chicago, and Kansas City (mostly to
rural customers) and through the company's retail stores (mostly to customers
living in nearby urban areas).

374

selves. At least 20 companies made part of their sales, and seven
made all of them, in this way. American Houses was a good example
of a company with a skilled central sales staff on the lookout for
large project business; and the producers of precast concrete houses,
for example, Vacuum Concrete, almost had to sell to large projects
because their system of construction only then became economical.
Most prefabricators, however, felt that distribution of this sort tended
to be spasmodic and made it difficult to achieve the steady and pre
dictable flow of production which they needed for greatest efficiency
throughout the whole pattern of operations.

2. Erection by Manufacturer

Twenty-seven companies sold all their houses directly to the con
sumer and then erected the unit for him, while 30 additional com
panies handled part of their distribution in this way.

Ordinarily, distribution of this type was localized, with erection
in the immediate areas surrounding the factory. It was found not
profitable to send erection crews several hundred miles in order to
erect one house or a small group of houses,20 and labor unions tended
to look with disfavor upon the arrival in the community of erection
crews who were not members of the union local.

An important trend during this period of boom housing was the
entrance into the prefabrication field of many large lumber dealers,
who preferred to do this work themselves rather than continue to
finance and supply builders as had been done before the war. Of the
companies interviewed, six were lumber companies which decided
to prefabricate and erect their own production. There undoubtedly
were many other lumber dealers not included in this survey who pre
fabricated and erected houses on a local scale.

A more dramatic performance was the erection by factory crews
of large site projects. Maximum economies could be obtained through
a combination of centralized factory production with mass erection
on a well-organized schedule on a large tract of land in the vicinity

20 At least part of the reason for the failure of Anchorage Homes lay in its
attempt to market its entire output (goal: 16 houses per day) direct to pur
chasers and to erect the houses above foundation with its own erection crews,
who often commuted to and from scattered sites several hundreds of miles from
the plant and were always hard to supervise.

375

of the plant.21 Well-publicized examples of the combination of plant
and site organization were the Byrne Organization and Kaiser Com
munity Homes, where whole neighborhoods were involved and site
location and planning became an obviously important factor in the
success or failure.

A few companies put up their own houses in group projects before
sales had been made to the ultimate purchasers. One reason for this
was to insure steady production at the plant; this channel of distribu
tion sometimes became, therefore, an adjunct to the more orthodox
methods of sales. If the prefabricator had not sold as many house
packages as were necessary to maintain a steady production rate, he
would build a group on speculation. Hamill and Jones had 100 such
houses under erection when interviewed. The California Prefab
Corp. was putting out four houses a day for its own erection, but
expected to sell only about two houses a week in response to orders
from outside customers. Naturally, operations of this sort depended
upon a continuing demand for housing in the price range offered, and
they were found only in areas of great housing shortage.

Where houses were largely plant fabricated, as were the Prenco,
Prefabricated Products, and Acorn houses, companies expressed a
preference for carrying out a good part of the erection of nearby
houses themselves, feeling that the combination of familiarity with
their product and ability to shorten an already brief site labor require
ment would be to their benefit.

B. Factory to Dealer to Consumer

* Most prefabricators considered a middleman between the factory
and the consumer a distributive advantage. The middlemen chosen
were usually independent dealers whose job was to relieve the manu
facturer of most of the marketing task, and to make prefabricated
houses readily available to more people than could the manufacturer
himself.

Certainly, in the eyes of the prefabricator, the principal function of
dealers was to enlarge the market for his houses. The prefabricator

21 Perhaps the clearest, although hardly a typical, example was that of the
Parsons Construction Company, in Canada, which set up a demountable wood
fabricating plant at each final project site, complete with movable tracks on which
to roll finished panels to the section of the site in which the proper part of the
erection process was in progress.

376

was limited in his ability to cover intensively what he regarded as
his market area, but strategically located dealers could be on the spot
all the time. By combining the orders from all these dealers the
prefabricator could maintain steady production and concentrate his
efforts on improving factory techniques.

Another advantage of the independent dealer was his familiarity
with the local market. He was better able to know when a member
of the community might become an active prospect for a house, and,
in addition, his community tended to regard him as "one of them"
and sales resistance therefore was likely to be less. He was also likely
to be of great assistance to the purchaser in dealing with local build
ing codes, tax regulations, finance problems, and site selection.
Furthermore, he provided the prefabricator with much-needed finan
cial aid. When the prefabricator delivered a house package, he was
paid by the dealer. Thus, the manufacturer had less money tied up
in the distribution process, and could devote more of his working
capital to production purposes.

There were many different types of dealers, whose function with
regard to the erection of their houses varied widely; they might them
selves undertake the erection, or it might be done by the producer
or even by the purchaser.

1. Erection by Manufacturer

Department stores occasionally were used as dealers for prefabri
cated houses. In such cases the house, a section of it, or a large-scale
model of it was erected in the store, thus offering to the store's large
clientele the opportunity of a detailed personal inspection with a
minimum of effort. A sales representative was almost always present
to describe the features of the house, arrange the sale, assist in financ
ing, and perhaps suggest tie-in package sales through the store for
furnishing the house. The factory delivered the house and erected it
on the site. Precision-Built Homes planned to sell in this way, utiliz
ing a "Precision Builder" who operated within a 50-mile radius, erect
ing all houses sold by the store in that area.

Only a few members of the industry believed that department stores
were likely to become important dealers in houses. Most felt that the
high unit value and low turnover rate of houses were not in char
acter with most other items for sale in such stores, and that customer

377

buying motives would not be likely to lead them there to buy houses.
Furthermore, it was pointed out that the many facilitating and fol
low-up activities connected with the sale of a house would be unduly
burdensome for high-volume fast-turnover department stores to as
sume. In short, prefabricators tended to feel that the most valuable
service the department store could offer would be to display the house
to a large number of people, and to design related furniture and
furnishing package sales which would ease the effort and expense of
furnishing the house. Thus, Adirondack Log Cabin and Anchorage
Homes, among others, made use of department stores only to display
their house models; the store did not enter into the sales transaction.

Use was made by 24 companies of general dealers in prefabricated
homes, with the understanding that the company would perform the
erection of all houses sold by these dealers. By and large, the pre
fabricators felt that more sales were made through these dealers than
through department stores, since their primary business was to sell
houses and they would be able to seek out prospective customers
more actively and to pay more careful attention to their needs.

In some cases the general dealers also provided, or arranged for, the
land on which the houses were to be erected by factory crews. The
Brice Realty Company, acting as an agent for Prenco, on one occasion
sold both a group of 230 houses and the land on which Prenco was
to place them. Of the companies using general dealers, seven dis
tributed almost exclusively in this fashion.

2. Erection by Dealer

More prefabricators elected to sell through dealer-erectors than
through any other distribution channel; 45 companies used dealer-
erectors in part, and 25 companies used them exclusively. For 27
companies from which detailed information was obtained, the average
number of dealer-erectors was 43.22 The PHMI in 1945 gave the
dealer-erector almost official standing as the preferred type of dealer
outlet in a resolution which recognized "the basic concept of selling
standardized, brand name homes, mass-produced, nationally adver
tised and mass-distributed to the mass market through dealers whose
functions will include sales, erection, servicing, and mortgage financ-

22 Of these, one claimed to have 400, two to have 100, and the rest fewer.
Five had fewer than five.

378

ing."23 Many of the companies were willing to indicate the fields
from which they drew their dealer-erectors, and this information is
presented briefly here:

35 Contractors or operative builders exclusively

19 Mostly contractors

15 Mostly operative builders

10 Real estate brokers or subdividers

7 "Financially responsible parties"

6 Lumber yards

3 Primarily selling organizations

As might be expected, the large majority had a background of some
kind of building.

The reasons for the popularity of dealer-erectors are worth investi
gating. First of all, as dealers, they enlarged the market area which
had to be larger than the immediate area of the factory in order to
maintain steady production. In fact it has been pointed out that
many experienced prefabricators felt they would have to produce sev
eral thousand houses regularly each year in order to attain the full
economies of industrialization. A well-organized chain of dealer-
erectors was believed to be the most likely way to reach such sales
volumes, and it could smooth out the irregularities in orders by cover
ing a variety of areas which the prefabricator otherwise might have
to neglect. The manufacturer was also relieved of the responsibility
of handling the mass of essentially local problems faced in erecting
the house, once the package had been sold. With a well-trained
dealer organization putting sales on an efficient basis, the manufac
turer could concentrate on production.

Dealer-erectors, while collectively enlarging the market area, were
able individually to concentrate efforts within their own relatively
limited market areas. Crews did not need to travel far to the sites;
factory-method advantages and erection economies could be mastered;
site expenses could be held down. Furthermore, dealer-erectors used
local labor, which induced greater local cooperation than would the
importation of factory erection crews.

To the ultimate consumer, the dealer-erector was a means of avoid
ing burdensome problems. Almost the only action needed on the
part of the prospective buyer was to sign his name to the sales con
tract and furnish evidence of being a reliable credit risk to the financ
ing agency. The major part of the prefabricated housing industry

23 Prefabricated Homes, 6 (December 1945), 12, reporting the winter meet
ing of the PHMI in Tulsa, Oklahoma, held December 3 and 4, 1945. Of course,
not all members of PHMI distribute in this way.

379

felt that this was the best way to get houses to customers at the
lowest cost, and many indicated that they had evidence that it was
cheaper than direct distribution from factory to consumer.

If a manufacturer distributed directly to the consumer, he had to
keep a sizable sales force in the field for high sales volume, and then
sales expense and commissions rapidly built up his operating expense.
In addition, a great deal of additional capital was needed to carry the
house packages from the time they left the assembly line until they
finally were taken over by the consumer. If the consumer was to erect
his house, the house package had to sell at a very low price; if the
factory handled the erection, further operating expense was incurred.
Well-trained dealer-erectors, making use of efficient selling techniques,
cut overall selling costs and were eminently qualified to handle the
sizable site-construction job involved in the average prefabricated
house. Planning ahead, they could pour foundations in warm weather
and thus continue to build houses in the winter season; this would
have the obviously beneficial effect of smoothing out the seasonal
variations in factory production.24

There were, however, some difficulties with dealer-erectors. Most
of them had previously been builders, used to working according to
local conventions and with local men. As a rule, they were rugged
individualists and good builders; they sometimes regarded new erec
tion techniques with disfavor. It was often difficult to persuade
dealer-erectors to take a limited profit per unit, on the theory that
they would sell many more units, in a seller's market and a period of
shortages when many prefabricators found it difficult to deliver the
promised volume. Some dealer-erectors were reluctant to tie up their
capital in foundations laid in anticipation of inclement weather.
Others, not wishing to displease local associates of long standing,
tended to buy less than the whole house package, omitting the parts
they would prefer to purchase locally. There was at the time of the
survey little real stimulus for the dealer-erector to build the sort of
alert service organization which prefabricators considered important
as a device to take care of minor difficulties once the house has been
built.

It was expected that most of these difficulties would be resolved in
a stabilized market, but nevertheless many of the large prefabricators

24 A good example was offered by National Homes, which even in the shortage
winter of 1946-1947 was making binding commitments with its more than 100
dealers for three months in advance, and which produced and shipped at a
steady rate of never fewer than 2^ units per day. So predictable a volume made
possible obvious procurement and production economies.

380

felt that they would have to develop a new, young, and flexible type
of dealer-erector.25 It was conceded that this scheme would be likely
to succeed in the degree that the conventional site work required in
the erection of prefabricated houses could be reduced.

Another very common middleman was the lumber dealer who, al
though not quite like the dealer-erector, yet fitted more closely into
this category than any other since he was a dealer who often pro
vided erection service. At least six companies used lumber dealers
to handle their house packages. Such dealers were willing to sell
the house either directly to the ultimate consumer or to the con
tractor. If the sale was made to the consumer, the dealer would
sometimes arrange to have a contractor with whom he had a work
ing agreement erect the house. In other cases he would erect the
house himself, in the manner of the typical dealer-erector. Some
times, however, the lumber dealer would sell house packages in
groups to contractors who intended to erect houses as a speculative
investment on land they controlled. In such instances, the lumber
dealer was in effect a distributor.

In some cases, for example General Houses (at one stage) or Pre
cision Homes, the preference for lumber dealers reflected the inten
tion of the prefabricator to market modular panels almost in the
manner of stock building materials. In other cases, for example
Peerless Housing Company, the prefabrication system was in a sense
a lumber-selling scheme; and in still others lumber dealers were used
in an effort to stay in the good graces of conventional builders.
HomeOla, on the other hand, used them because the company
marketed by carload lots of five houses, and lumber dealers were
able to handle this quantity on a single order.

Whether serving as dealer-erectors or as distributors, the lumber
dealers tended to handle houses simply as a side line, and they tended
not to make much of a sales effort. Furthermore, they usually were
not well equipped to provide many of the specialized services which
should accompany the sale of houses. HomeOla estimated that "re
tail lumber dealers accounted for the sale of only 2,000 out of the
known total of 37,400 prefabricated houses sold in 1946," 20 and the
company eventually decided to concentrate on other channels of
distribution.

25 Both American Houses and National Homes, for example, were planning
to take well-trained college graduates into their organization for grooming, and
then to finance their debuts as dealer-erectors.

26 The HomeOla Dealer Info-Service, Bulletin M-18 (June 9, 1947), p. 3.

381

3. Erection by Purchaser

The quantity of sales made by dealers under the terms of which
erection was left as the responsibility of the consumer was relatively
unimportant. Department stores made a few sales in this manner,
but without great success. General dealers operating in rural areas
probably sold the most houses to be consumer erected, although
Hamill and Jones, which had previously sold direct to consumers,
was an example of a company building up a dealer organization but
still making sales in some areas for purchaser erection. Pre-Fab In
dustries had rural dealers scattered over the states of Indiana, Illi
nois, Ohio, Michigan, and southern Wisconsin who made sales pri
marily to hatcheries and other "barnyard

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "The prefabrication of houses"

See other formats